Anna Loverus - Internet, digital strategy, and marketing

Social Media Risks and How to Protect Yourself Online

Anna Loverus on March 20, 2018

With all the information we give the social networks, the more information they have about us, to hold and use as they please. People are becoming more aware of social media risks, but what risks are there?

We are not just uploading more of our lives on the internet; we’re also spending more and more of our days consuming, sharing and commenting on social media platforms. Naturally, risks related to social media are increasing. The goal of this post is not to make you stop using social media, or the Internet for that matter. Being smart online, especially if it’s part of your job, is only possible if you know about the risks and how to limit them.

Hacked Social Media Accounts

Getting your Social Media accounts hacked is increasingly common and probably one of the most significant social media risks. It happens to both large corporations (HBO, PlayStation) and people with small personal accounts. A hacker is anyone from an ex-partner who can guess your password or a previous employee who made a copy of the password lists before they left, to professionals with more advanced methods.

But how significant is the risk? Facebook said in 2011 that 0,06% of all logins on the platform was “compromised”, and during that time the platform had about 1 billion logins each day, so that is 600 000 problematic logins every day. Google reports that 20% of social accounts are compromised “at some point”. I would never have guessed that 1 in 5 accounts would be compromised, that’s a lot of sad users.

What can you do to prevent it?

Use secure passwords and different passwords on every service you use. I recommend using 1password to make your life with secure passwords less of a hassle.
Use two-factor authentication to make sure that it’s harder for someone to log in even if they have your password.
Avoid browsing (especially logging in to) social media accounts on public wi-fi, the techniques for snooping around and waiting your login credentials are getting more advanced.
Never give account or page credentials to someone who contacts you directly, not even people who say they work with customer service for the social media itself, or that they are your colleagues.
Never download apps, especially not apps that want’s give permissions to post on your behalf.

Social Profiling

Social Profiling is when someone uses the information on your social media accounts to create an opinion about you or “measure you” based on your interactions or influence on social media. This is typical behaviour today, both from employers and the people you’re dating.

Around 43 percent of businesses used information online when they’re decided not to hire someone. And 40 percent use social media when they screen candidates. Sure, this is a significant number, but your life online shows only a small part of you as a person. And if you don’t have a social media profile, or if your “influence score” is low, apparently you’re less worthy of an employer.

What can you do about it

It is hard to do anything about the behaviour of potential employers and Tinder dates. However, you can make sure to know what those searches will say about you, and you can adapt what comes up.

Use the incognito search mode to search for yourself online to see what comes up.
Remove pictures that are sketchy, strange or that someone can use in an entirely different context
Use the “view as” setting on Facebook, to understand how your profile looks to the public. Make sure that potential employers can see only the best information about you.

Cyberstalking

Although cyberstalking might feel as bad as any stalking, it’s a bit different when it happens online. It’s not uncommon that it’s combined with offline and real-time stalking as well.

There are some differences between “traditional” stalking and cyberstalking. While traditional stalking often happens to women, cyberstalking is affecting men 40 percent of the time. One other difference is the stalker. Often, it’s an ex-partner or someone with a connection to the victim doing “traditional” stalking, while cyberstalking is widely done by complete strangers.

The fact that it is so easy to collect information about someone online is probably behind the significant increase in cyberstalking. A cyberstalker can use information about your geolocation, and this is automatically turned on for most smartphones. And it is rather easy to follow your life and regularly see where you are if you use apps like Swarm, check in on Facebook or tag your photos with Instagram.

What can you do about it?

Turn off the automatic use of geolocations.
Don’t actively check in with Facebook, geotag your images on Instagram or use apps like Swarm
Limit the information you share on Facebook and other platforms so that it is only available for your friends

Third party information sharing

Many websites add cookies to your device when you visit them. Most cookies role is to find returning users to improve your online experience or keep track of users to improve their service. The problem is, however, that some cookies can crawl and follow you on other websites. Therefore, they can get a detailed view of your online behaviours.

Many services you opt-in to use, like Facebook and Google, record a rather significant amount of what you do. They are tracking your activity when you use their service online. This is everything from messages and comments, what you share, and for how long you looked at an individual piece of content.

They can then sell all this data about you to either another company or as part of their advertising services. You will, unfortunately, have microscopic say about who’s buying your information and for what causes they can use the data.

What can you do about it?

Take a look at your Facebook Ad Preferences. Remove the things they’ve collected about you that you don’t want Facebook to use when they show you ads.
Opt out of Facebook’s tracking and ads program. Choose “no” for both “Ads based on my use of websites and apps” and “Ads on apps and websites off of the Facebook Companies.”
Opt out from being tracked by companies in the Digital Advertising Alliance. To opt out from all the participating companies, click on “Choose all companies.”
Install Privacy Badger from the Electronic Frontier Foundation (EFF). “Privacy Badger is a browser add-on that stops advertisers and other third-party trackers from secretly tracking where you go and what pages you look at on the web.”

Warrantless searches

Warrantless searches are searches performed without a court-ordered “search warrant” and is the most common type of search conducted by law enforcement. According to Katz v. United States (389 U.S. 347), courts determined in 1967 (long before the age of social media) that if a person willingly makes information public is not protected by the Fourth Amendment and, therefore, it requires no warrant.

Information such as direct messages that you want to keep private is often interpreted by the law as not being “private” since there is no guarantee that the person receiving the message will keep it secret. The NSA has also worked with Facebook, Google, Apple and other tech companies through the Prism program. Part of this is sharing of user data such as emails, search history, instant messages and transferred files.

What can you do about it?

It is possible to do warrantless searches and record information since all our online activities are out in the open. The best thing to do is, therefore, to “hide” your actions.

Use a VPN (preferably with double encryption and “no logs” policy) on both your phone, your computer and all other devices
Use an end-to-end encrypted application for instant messages
Meet up in person when you have to discuss sensitive information

Social media risks, some last thoughts

Make sure to check that you have decent security related to your social media. It might not seem like a significant risk to you now, but keeping certain things out of the public eye (or the Google Search) is not stupid, it’s smart.

The costs for getting your accounts compromised are pretty high, and it is not easy to get compromised accounts back if they get hacked. Most of the proactive suggestions in this article, to cut your social media risks, is probably possible to fix. That is perhaps time very well spent.

Two-factor authentication for beginners

Anna Loverus on March 14, 2018

Databases will always get hacked. Passwords will sometimes get into the wrong hands. You should, therefore, make sure to have a system that reduces the damage a leaked password can create. Two-factor authentication is a way to decrease the risk related to living life online.

I’ve realised that few people working with the internet and its services as primary business tools do not know about this technology. Even fewer people use it. That is why you are now reading this blog post…

How does two-factor authentication work?

Two-factor authentication adds an extra layer of security on top of your username and password combination. It makes it harder for a hacker to get access to your account even if they know (or break) your password. For someone to get access to your account they need either your phone or some other “key” to log in. The service use the phone or key to verify that the person logging in is the legitimate owner of the account.

When you log in to an account it will ask for a code you get from an app in your phone, or through text message. Some prefer to add a USB-key with a button that you press to authenticate the login. There are multiple solutions and you should pick what feels easiest for you.

It’s actually the same solutions that many banks use. They give you a physical card with codes, or a small digital device providing you with codes when you press some numbers. You need your login details and your specific code, to use your bank online or through the phone. This is to make sure the bank know that you are you.

When should you use two-factor authentication?

You should basically activate two-factor authentication, 2FA, for every service where it’s possible. To find out if a service offers two-factor authentication you can use a service called Two Factor Auth (2FA). It lists most online services and links to the set-up documentation for each one of them.

You should activate two-factor authentication for at least: E-mail, Facebook, Twitter, Dropbox. All places where you have sensitive personal information, such as your bank and health-related accounts, should have 2FA activated and everything work-related should also have maximum security. When it comes to work accounts, the damage would be large, not just for you, but also for your employer.

But it’s such a hassle…

The hassle setting up two-factor authentication is worth it. Just compare it to the trouble you will have to go through if someone finds their way into your accounts. Sometimes it’s hard just to take down a hacked social media account. And if someone has access to your email they can easily reset other passwords creating, even more, damage.

I was once able to prevent a login attempt from somewhere in Asia because I had two-factor authentication. Someone knew my password and I got an e-mail about the login attempt but since they didn’t have my phone they couldn’t access the account. Instead, I could log in and change my password, keeping them out of my account.

Anything else?

Well, while we’re at it: You should also use a password manager, like 1Password, to make sure all your passwords are secure, i.e. long, complicated and used only for one account. It’s a pain to move into a password manager and change all your passwords, but you should do it TODAY. It further reduces the risk of some stranger getting access to your account.

Why you should take a break from Facebook marketing

Anna Loverus on March 9, 2018

Marketing is at times a bit like finance. You place a bet on what you think will perform, and then you wait and see. Luck plays a part in the result, but often experience adds to the success rate.

But just like an investor is not placing bets by guessing, neither are marketers. A variety of information and some intuition goes into every bet. We often base our bets on a combination of historical data, the calculated risk and the potential return.

Both marketers and investors are looking for the best return on their investments. For every dollar we spend, we want as much back as possible. For an investor, the gain is often in dollars, but for a marketer, it might be in sales growth or recognition. And while it might be easier to calculate ROI for a financial investment, the underlying mechanism is very similar.

Betting on the right thing

Marketers, just like investors, need to know how their surroundings are changing to make sure they place their bet strategically. Today, the surroundings for marketers primarily consists of algorithms: Facebook’s news feed, Google search, YouTube’s video recommendations, and some others.

When algorithms change, a bet that once was lucrative won’t pay off anymore. And while investors are quickly moving from an investment that is no longer paying off, marketers seem to keep going for a long time before they realise they should change something.

Why is that? Why do so many marketers continue with efforts that do not pay off? Is it because we are stupid? Or lazy? Is it because marketers don’t follow what’s happening in their field as closely as investors read the financial press? I don’t have an answer, but it fascinates me greatly.

Facebook marketing and ROI

Last summer the debate intensified around “fake news” and Facebook’s part in the outcome of the US presidential election. Quickly after this, the referral traffic from Facebook started to dive. The Facebook algorithm changes from time to time, but they didn’t say anything about changes at this time.

Facebook has made changes before as well. The platform has moved from being a source of free engagement and referral traffic, to become an advertising platform where you buy your impressions. These changes make a lot of sense; Facebook is a for-profit company.

As you see, betting on Facebook for marketing results comes with a high risk. If they change things on their platform, you will lose. Even if you’ve invested all your money on the Facebook platform for years now, you still might end up with nothing at the end of the day. Since you only “borrow” the relationship with your fans from Facebook, there is no real value in a Facebook page even if it has hundred thousand followers.

During 2018 we have continued to see multiple indicators suggesting the ROI on Facebook marketing is declining. When you, as a brand, publish content on Facebook today almost nothing happens. Some features, like groups and live streams, are ways for brands to still be relevant for users on the platform, but it’s very hard to get a good return on your investments.

Maybe, it’s time to take a break? At least, if you don’t have an advertising budget and if you spend a lot of time on your Facebook content and still get small results.

What to do instead of Facebook marketing

A lot of people think that digital marketing equals Facebook marketing. Some might throw Instagram and Twitter into the mix too. And while Social Media marketing is an essential part of digital marketing, one skill you should have in your toolbox, it’s not everything.

You need to move over to platforms where you have more control of your success. Brands doing content marketing should most likely focus on other platforms, and increase their focus on SEO and e-mail.

E-mail marketing and blogging

E-mail is an alternative to Facebook because your success is very closely correlated with factors you can impact. Search is another (now) stable platform, although not entirely without risk since it’s still an algorithm behind it.

Business blogging have for long been a bit uncool, “why should anyone wanna read a business blog”, but I think it is will soon have a revival. It’s safer than Facebook if you want to get back what you invest. A bonus is that more formats and topics will work in a blog setting than on Facebook. To be successful, you don’t have to do 45-second videos with text in the frame.

Podcasts and YouTube

Podcasts and YouTube will also become more popular when Facebook investments are no longer creating great returns. YouTube is apparently the second largest search-engine in the world, after Google, so there’s potentially an even more significant return to make from this platform. Podcasts are unique because users can use multiple technical solutions to consume your content, something that almost makes it similar to e-mail and reduces the investment risk.

Not that many marketers are very good at working with podcasts or YouTube today unless it is one-offs or sponsorships. Much of the best material is instead generated by “Podcasters” or “Youtubers”, with the content as their core product. But both these platforms now feels more stable than many other content platforms (as long as you are not trying to use it to make a living and don’t have to care about the compensations models), and the potential audiences are large.

How I try to place my bets

Personally, I began to refocus my Facebook marketing initiatives for most clients or projects last summer. More than a year ago (maybe earlier) I moved away from Twitter and Snapchat and recommended others to do it too. The platforms weren’t delivering enough results, and the risk of them just disappearing felt too big to for the small ROI they produced.

Instead, I’ve started this blog, and I have an e-mail newsletter (you can subscribe in the sidebar to the right). I’m also looking at starting a pod or a YouTube channel when I have the time. Sure, this tiny blog is not Spotify, but I would probably recommend them too to move away from Facebook at the moment.

My focus is on consistency and not quantity. It’s often wise not to hurry when it comes to scaling content initiatives; it takes time to build an audience.

My tools for Digital Marketing – Data Processing and Data Analysis

Anna Loverus on March 5, 2018

I often get questions about tools for digital marketing. Great topic for a blog post I thought, but when I started to write it became so massive that I realised I need to make a blog series out of this topic. In this post, I’ll walk you through the tools I use for data processing and data analysis.

Do you need technology to do marketing? Well, not long ago you could get by pretty well without it, but marketing is becoming more and more about technology. Sure, the underlying idea of understanding people is mostly the same, but you will soon be left behind if you don’t add technology to your toolkit.

Finding tools for digital marketing

With a background as a tech journalist, I have this weird interest in new technology. Whenever I find something new I can try out, I get excited, and I love spending time on ProductHunt to see what’s new. Maybe not the most normal thing to do on a Sunday night.

I’m always on a hunt for two types of tools: 1. technology that can simplify or even automate part of my current job, or 2. technology that can give me possibilities that I don’t have today.

But I started out like most people: writing reports, making presentations and sometimes using spreadsheets to do calculations. I guess most of my time is still spent putting thoughts on paper, and my every day “martech stack” is still pretty basic.

Everyone can learn data analysis

Since I’m going to talk about tools for data, I just want to say that my background is not in statistics or engineering. I do have some university credits in basic statistics, and I once knew how to perform a significance test in SPSS. But I’ve learned most of my marketing data skills by doing. I started testing with small side projects, spending time with both MOOCs and tutorials online has been a pretty good way for me to learn.

But I would also say that part of why I learned it all was that no one else around me knew how to do data analysis, so if I didn’t try to figure it out on my own, I wouldn’t have any quantitative insight at all. Not using the data I had access to in some way felt more stupid than to try to do some data analysis on my own.

My tools for marketing analytics and data analysis

I regularly walk clients over slides with data analysis nowadays. And I do everything from the first export to the visualisation on my own. I think I have three types of reoccurring projects where I need my data skill set:

Auditing – Looking at historical data, in a delimited and pre-defined context, to find how something performed
Monitoring and Measuring – Visualising data in real-time, creating opportunities for better decision-making
Research – Looking at trends and decoding information in an unknown context

The tools I use are either for processing, analysis, or visualising and presenting data. And once again, it wasn’t long ago I didn’t know how to split up a CSV-file without help from Google; Hence, you shouldn’t be intimidated or feel like it’s something you cannot do yourself.

Tools for Data Processing

Google Sheets

Google Sheets is a great way to work with certain types of data, especially medium-sized data structured in rows and columns. My first real relationship with spreadsheets started here, and its interface made me feel safe(r than other spreadsheet tools). I also realised I could find the answers to most of my Google Spreadsheet related questions online.

The first time I used Google’s spreadsheets was for personal budgets and other types of simple calculations. But I like tracking stuff, so most of my courses at Uni had a spreadsheet with all the tasks and deadlines and suggested readings, and I kept track of my progress using colours and conditional formatting. I used it more to create well-structured files than to process data or calculate anything. I later started to do “Content Calendars” for clients in Google Spreadsheets, since it was easy to get an overview.

My introduction to using Google Sheets

But my first real use of Google Sheets for work was to create monthly performance reports. I exported standard data files from Facebook, Twitter, Pinterest and the other sources I needed. In my Spreadsheet I had one “master sheet” calculating all my KPIs, referring to data from sheets named per data source. When I wanted to update my KPIs, I just overwrote the data in my import sheets, and all calculations in my master sheet refreshed automatically – as long as all the columns in my export were in the same place, but usually, they were.

I later started to write Google Script to make more automatic data imports, not having to download and upload files. Using scripts to get access to data from an API makes it possible to have automatically updated data in your sheets. This method makes it possible to use Google Sheets as a simple dashboard solution if you need to. I also use tools like Supermetrics to automatically import data from different APIs through a Google Sheets plugin (more on that further down).

Microsoft Excel

While more advanced users might disagree, I would say that Google Sheets and Excel are very similar. You can do most things on a fundamental level in both tools. Some things are more comfortable to do in one than the other, and I guess that’s why I continue to use them both simultaneously.

I tried to stay as far away as possible from Excel for very long. To be honest, it felt like Excel was something only dull people needed, and my biggest fear in life was to seem boring. So, I stayed far away. For a very long time, I opened my .xlsx files in Google Sheets.

When is Excel better than Google Sheets?

The more I work with data, the more time I spend cleaning and prepping data files for data analysis. Additionally, my data files got heavier (and still do). So, the first feature I needed in Excel for was to use it without an internet connection. I know you can work with Google Sheets in offline mode too, but it doesn’t feel as safe.

Quite quickly I also learned that Excel was a bit more stable than Google Sheets when I was working with larger files. (It’s not like Excel is perfectly durable though, I’m very often looking at the spinning rainbow ball when I work with Excel). One elegant Excel feature is that you can turn off automatic calculation and make all your changes to the document before it calculates what’s in your cell. This feature might give you some extra processor power when you need it.

Today I often use more advanced functions in Excel than I did in the beginning, and I could probably do in Google Sheets with a plugin, but it’s neat that they are already part of Excel. The type of data imports I do in Google Sheets, using Google Script is not something I do in Excel, even though it’s possible. And as soon as I need to share my documents with someone else, working in Google Sheets is often much more accessible.

Atom

Atom is a text editor (like TextEdit on a Mac or Notes on a PC). But why do you need a text editor to work with data you might think, it doesn’t make any sense? Well, at times when I work with data processing, I get across data stored in JSON or XML-files. When you try to read what’s in there, it looks like gibberish. A good text editor is very helpful when decoding these files.

But any text editor won’t help you; you need a good enough one (like Atom or Sublime) so that it formats the text in the file based on its language to make it more readable. This process is often called prettifying, unfolding code that is hard to read into a structure that is friendly and easy to understand.

For example, Facebook Ads API display the targeting data per ad as a JSON-object. If you try to read it in a spreadsheet column, you will struggle. But Atom formats JSON beautifully if you activate one of the “prettyfiers” that comes with the tool. You create a JSON-file, paste your FB-targeting data into it, and save it to your hard drive. On the save it will magically become formatted and (almost) readable. And the same is true for many other file formats that you might need to decode.

Tools for Data Analysis

Tableau

I felt intimidated when I first found Tableau. Partly because I didn’t already know about the tool or that there’s a whole field called Business Intelligence (where people are doing data analysis for a living). But also because it felt like I needed an exam to have the right to use it.

Business Intelligence software makes data analysis much more manageable, so I decided not to care about my lack of previous knowledge and downloaded a demo version. But I had quickly found myself with two problems: 1. It’s hard to get started with Tableau as a beginner, 2. Tableau is super expensive.

When should you use Tableau?

I use Tableau to extract information from my data, to look for learnings or insights that I could never see without help from software. For instance, are we spending our advertising money on the content our clients engage with or the material we think is best? How are different segments of our users behaving when they interact with our product? I ask questions to my data through Tableau, and most of the time it shows me that my first gut feeling is incorrect. But at the same time, Tableau often shows me things that I had no clue about and would have never found through empirical studies.

Tableau is hard, but this is just because it’s different from most other software you’ve used before, so we cannot translate much earlier knowledge into this tool. Your data are categorised in dimensions and metrics, and in the beginning, nothing makes sense.

Teaching myself Tableau is probably the best thing I’ve done – it made me understand data on a deeper level. I’ve always known about the difference between boolean variables, strings and integers from my background in programming, but it became much more tangible when I started to look at different data sets, trying to extract information from them.

Tableau has excellent tutorials online, so you are not alone in this process of not having a clue. There is also a great forum where you can read and post. Users always help each other out which is nice. But Tableau is not (yet) as well documented online as Google Sheets and Excel. When you have a problem, you might not find the answer on your first search. Stack Overflow is another place if you need help.

R

Last but not least, R is a tool that I don’t have to use that often. It is a useful backup tool for analysing data, and it can solve some problems that are not possible with the earlier ones I’ve talked about in this post. If you have a coding background, you will have fun learning R. But if you don’t R will seem pretty hardcore since you are interacting with the program through writing code and not through nice visual interfaces. It is similar to the terminal on your computer, in many ways.

I only use R when I have extensive data files, or if I need to calculate relationships between data points or data sets. Neither of these needs appears very often in my everyday life in marketing. But I did a network analysis once, that wasn’t possible to do in any of the other tools. One key feature is that you can run your calculations on a server in the cloud if your computer cannot handle the size of your data sets or calculations. I’m not saying you should start doing that, but it’s good to know that it’s possible.

Overlaps between data sets are also hard for the other tools to handle, but R nicely calculates differences and draws Venn diagrams. R can do a lot of graphs and visualisation, but they are not the most visually appealing, so I don’t recommend to use it only for that.

How to learn R

If you think you’d like to learn R, multiple MOOCs can help you get started. I’d recommend taking one of them. But if I were you, I’d start with one of the other tools; I only use a fraction of the functionalities in R. But sure, I plan on getting better at it, I just have some other things in life that I might prioritize before that…

The difference between an offline and online audience

Anna Loverus on February 28, 2018

The one thing I get most requests about these days is helping out with online audiences that are not working. And I realise I see the same problems over and over. Naturally, that’s when a blog post is born.

Most people working with marketing audiences today started long before online marketing was the norm. They still create audiences for the online world like they did (or still do) for the offline world. But the two are very different, and naturally, audiences don’t translate very well between the two contexts.

The difference between online and offline audiences is how you decide if someone is part of your audience or not.

Your audience hypothesis

Before you create an audience for your ads, you usually have an idea of who you want the reach with your product. If you’re going to market a contraception app, spending your marketing budget on women between the age of 23-45 seems fair, but if you try to sell fancy cheese, your audience is somewhat different and probably should consist of cheese lovers with enough income to spend on cheese.

Sidenote: Some people work with personas to get to know their audiences. I try to avoid that since I find it limiting. I will save my take on personas for a separate post. But its safe to say that personas create a lot of trouble when people are trying to reach their “personas” with online ads.

Let’s get back on track. Often you have to translate your business audience into an advertising audience. Maybe because you want to personalise your ads based on preferences, or age; Or, because you have a small budget and want to make sure you spend it on those who are most likely to consume it.

The limit with offline audiences

In the offline world, you have very little information about people. You often know the average income level in a zip code or a magazines rate of female readers, but you don’t have rich profiles or detailed information about a single person.

So, when you advertise offline, you do it in a zip code where the average income level is similar to what you think your audience earn. Or you choose a magazine with mostly female readers in a relevant age span. But you will never know if they like cheese or are trying to get pregnant.

What is a “proxy”?

In statistics (yes, building audiences is statistics), a proxy variable is a variable that is not in itself directly relevant, but that serves in place of an unobservable or immeasurable variable (Wikipedia).

The zip code is a proxy for income, and you can use a women’s magazine as a proxy for gender and maybe also specific interests if the magazine focuses on a particular type of content. Marketers try to find good proxies to make their advertising do better – but it is hard for some products and services.

Say you run a house cleaning service. You believe you should try to reach women because they feel more responsible for house cleaning (bleh). But you also want them to make enough money to afford your service. But how can you narrow it down further? Say you want to reach women with demanding and high paying jobs. Then you can advertise in magazines that people in this category are likely to read.

This magazine sure seems like a good proxy for your audience, but you won’t know how many of the magazine readers who are relevant to you. Some women who read the magazine might have demanding and high-paying jobs, but they have husbands that do all the housework. And others might not have a high paying job yet but wish to have it one day, so they are reading the magazine as inspiration.

Offline audience spill-over

So among those you target with an offline audience, only some people are the ones you’re trying to reach. You will have “spill over” to other groups that you’re not looking for. It’s the same with “out of home” ads, you can put them up on the bus stops in an area where the income level is high, but that doesn’t mean a significant share of those walking by, looking at your ads, won’t have that high income.

The benefit of online audiences

When you create audiences online, you have much more data about people. Either, you have your own customer data (“first-party data”), or you use data from Facebook or Google (“secondary data”), or you buy data from someone else.

Facebook and Google collect data not only on their platforms but all around the web and in apps through different scripts such as Facebook like buttons and analytics scripts. This data collection is why they know so much about their users, and this is why their advertising solutions are thriving online.

When you have data directly about each person, you don’t need proxies. You can target users on Facebook directly on income level. Or cheese interest, or their interest in contraceptive apps. These targeting possibilities are why Facebook ads can become so relevant. We don’t need to guess if those interested in cleaning services are also reading a specific type of magazine, or are more likely to have a gym membership. We can target directly towards people interested in cleaning services.

Why your Facebook Audiences don’t work

Almost everyone that contacts me, about poorly performing online audiences, have built them incorrectly. Hence, they are using proxies instead of pinpointing the behaviours or interests they are trying to target. This set-up makes their ads taking a massive detour in deciding who is relevant and who isn’t.

And the use of proxy-based audiences for social media is widespread. I’m continually meeting both media agencies and social media marketers that are doing it daily. But it is just MUCH MORE WORK that will give you a WORSE RESULT. So it’s pretty easy for me to recommend you to stop.

Can marketing personalisation be unethical?

Anna Loverus on February 24, 2018

Personalisation is an ongoing marketing trend: far from new, far from over. Being specific and relevant to every single customer is a powerful marketing tactic, often appreciated by those receiving your message. But it has some downsides worth knowing.

I don’t know how many times a week I talk about the magic triad of marketing. Providing customers with 1. the right message, at 2. the right time, in 3. the right place. Personalised experiences give you tools to nail the first part, creating more relevant messaging.

We can personalise experiences in many ways. Personalisation online includes everything from adding customer names to welcome phrases in emails, to changing website content and design for every single visitor with help from artificial intelligence. The goal is often to increase customer engagement and conversions through improved relevance.

In this post, I talk about both targeting and personalisation. And why it’s not the same thing, doing one without the other is hard. If you have personalised content, you need to target the right person. If you target a specific group, you often (but not always) do it to become more relevant to this group. This will turn into personalisation if you narrow your audience enough and change the content to suit them.

The current state of personalisation

Few companies use full-scale personalisation today, but some display content on websites or in emails based on a customer’s previous behaviour. Others are creating content for different segments of customers, although there might be more than one person in each. This marketing tactic is underlying when customers with a birthday in November will get the same email, or those who bought notebooks recently will get the same ad.

Netflix or Amazon are both using advanced personalisation. They select every item on display specifically for you, and it sometimes feels like they know you better than you know yourself. Looking at my Netflix recommendations is like looking directly into my brain, it reveals every little quirk I’m not talking about in public.

Why personalisation needs careful thought

While personalisation is a powerful marketing tactic, it is sometimes perceived as “creepy” by customers. In the best of worlds, people like their personalised experiences. At times they might “only” get a bit uncomfortable. But personalised content can also be very unsuitable or even unethical.

There will be a constant battle between personalisation and privacy, and it is essential for marketers to know a bit about the risks. I will discuss two types of content personalisation and online targeting situations that are more problematic than we might think at first – but there are of course many more.

Many (most?) digital marketers use these methods today without knowing it is problematic. In this world of continuous consumer data collection, we need to discuss marketing tactics and marketing ethics – but we don’t. This blog post is far from a complete guide, but it might get your thoughts started.

Targeting based on health data

We give away a lot of health information online. Googling symptoms, looking for home cures, worrying about constant headaches or trying to break our bad habits. Our online behaviour additionally gives away many cues about our mental health – how we interact with social media, for instance.

But just because we can target an obese person with diet tips or a depressed person with advertisements for self-help books or therapists, doesn’t necessarily mean we should. And it just becomes even worse if you start targeting cancer patients or their relatives (for instance, people who have visited the cancer wing at your local hospital) with ads about funeral services.

Health issues are one of these things people don’t want others to know about, and we will find that targeted ads violate our integrity or are intrusive. While it is probably okay to communicate with parents around the pains of having sick kids in February, we should carefully make sure we don’t fall over.

Retargeting users with health-related content

Another issue that often appears is retargeting based on earlier shopping behaviours. But what if we spend some time looking at health advice, comparing medicine or googling back problems? If you have browsed around for self-help books, and all of a sudden get suggestions for all the other self-help books you should buy too. Or if you look for specific medical treatments or drugs and it follows you around the web for weeks, that is not very good marketing.

Personalised or retargeted content based on someone’s health, is often a terrible idea. It will give people the feeling that you know things should not know about them, and the ads likely will not perform since customers find this intrusive. Pick another marketing tactic.

Political opinion as targeting

Using someone’s opinions for targeting is an ancient (well, sure) marketing tactic. It is not only about how someone is voting; political views include so much more.

Many niche opinions that might seem harmless to you is controversial or even illegal in parts of the world. People engaged in gay rights, pro-drugs or anti-abortion movements – might end up in unpleasant or even dangerous situations if their opinions get out. Maybe not in your context, but in theirs.

With Facebook’s interest targeting, it was previously possible to target people with interest “Jew hater” or “How to Burn Jews”. And while Facebook took away these particular targeting options, many others are still there. You can target people based on ideology, such as liberal or conservative, or fans of a specific political party.

There are cases where targeted ads seem to have put people at risk, or in uncomfortable situations, both in online and offline contexts. For example, people do lose their jobs when their employers find their opinions inappropriate.

Sure, if you live in a democracy and everyone is free to speak their mind openly, the risk might be small. But it was not long ago that people had to hide their political views from friends and family in parts of Europe – and people still get into trouble because of what they think politically, all over the world.

Calculating the risk

Giving someone’s political views away by showing them personalised content can lead to consequences we might not consider, or even understand when we do the targeting. These risks vary from market to market, but you are responsible for deciding what is okay and what is not.

Often, when you do use someones political opinion for targeting, you are using it as a proxy to categorise people with a specific set of values, behaviours or traits. Using these traits directly is often more ethical, and it will make your marketing much more relevant to your customers.

(A side note: Using consumers political residence for targeting or personalised content does not mean you cannot promote political content. But who gets your political content and why needs to be an active and careful decision based on fair grounds.)

The need for marketing ethics will increase

Marketing ethics will become a big deal over the next few years. All marketers will need to know more about ethics than we do today, or we will get in trouble. Many companies will get lost, and they can end up in some real trouble putting their brands at risk.

Technology, both specific to marketing and not, will continue to give us marketing opportunities that are inappropriate and unethical. New tools help us target and personalise messaging, and they are continually getting better and more popular. But most marketers using these new tools have never had to think about what is okay and what is not. Instead, we do things just because we can, because that’s what we’ve previously been doing.

To follow the law will never be enough

Customers don’t mind to barter with their data if they get access to free valuable services in exchange. But they are picky about how brands use their data. As soon as ads feel intrusive or inappropriate, customers will not engage. This behaviour creates a paradox where marketers will have to balance between extremely relevant to each customer, and not to make customers feel uncomfortable with why they get what they get.

Policy and legislation will soon create more boundaries for digital marketers. But it won’t ever catch up with technology. Law development being slow is a problem for computer scientists and programmers as well. So, we all need to brush up our ethics.

I’m not saying that business has ever been entirely ethical. Too much money is often at stake. But marketing is easy to review since it’s on display, so I think we’ll have to get in line pretty quickly – or we will get in trouble.

Do you have any thoughts on this? Please leave a comment!

Därför ska du inte ta debatten med rasisterna på Facebook

Anna Loverus on November 29, 2017

Med ett val runt hörnet och polarisering som en identifierad samhällsutmaning är det många som funderar på hur man kan bidra till ett bättre debattklimat på nätet. Här är några konkreta förslag på vad du kan göra, och vad du borde undvika.

Ett återkommande tema under förra veckans Internetdagarna var att vi alla behöver ta individuellt ansvar för Internet. “Vi kan göra det tillsammans” är väl något av det minst farliga man kan föreslå, även om jag inte tror att vi kan bygga ett fungerande internet enbart på individuellt ansvar. Men det finns vissa saker som är bättre att göra än andra om du vill att trollen ska få mindre makt och hatet minska. Vissa saker, som kan verka oskyldiga, är rent av kontraproduktiva. Så om vi alla ska “göra det tillsammans” så gäller det att folk förstår vad som funkar och inte, utifrån hur internet ser ut idag.

Om målet är att åstadkomma skillnad, gäller det att jobba så smart som möjligt.

Facebook är en teknisk plattform med vissa spelregler. Att förstå hur saker och ting sprids på nätet, och vad som påverkar detta, är nyckeln för alla som försöker höra sin röst hörd. När det kommer till just främlingsfientligt och rasistiskt innehåll är en del av initiativen och idéerna om vad den enskilda borde göra för att stävja detta, rent kontraproduktiva. Min avsikt med den här texten är inte att klanka ner på dessa initiativ, utan att – utifrån mitt perspektiv som någon som dagligen jobbar med att utnyttja Facebooks mekanismer – hjälpa till att förklara vad som vore att föredra istället. Självklart ska vi inte blunda för främlingsfientligt innehåll på nätet. Inte heller för annat innehåll där vi inte håller med. Men om målet är att åstadkomma skillnad, gäller det att jobba så smart som möjligt.

Här kommer fyra punkter du med fördel ska göra istället för att “neutralisera” eller debattera i kommentarerna på främlingsfientliga, rasistiska eller i största allmänhet korkade inlägg på Facebook. Eller, den allra första punkten är att anmäla det som på riktigt är rasistiskt. Snart kommer även funktioner från Facebook som gör det möjligt att anmäla falska nyheter och misinformation. Men om man vill göra mer än så? Om man vill verka för att världen ska bli en bättre plats? Det är faktiskt lättare än du tror.

Vad kan du egentligen göra för att minska polariseringen på Facebook

1. Ignorera det som är kass

Kommentera inte på främlingsfientligt eller hatiskt innehåll för att “ta debatten”. Kommentarer, reaktioner och hur länge vi tittar på något påverkar direkt hur många andra som får se det eftersom Facebook anser att det här innehållet är relevant för många. Facebooks algoritm “prioriterar upp” innehåll som får mycket likes och kommentarer, så när du kommenterar på poster du inte gillar kommer det innehållet alltså att nå ännu fler.

Den goda tanken om att många gemensamt ska gå in och skriver neutrala eller debatterande kommentarer på rasistiskt och hatiskt innehåll är alltså fullkomligt kontraproduktiv och hjälper alltså istället till att sprida detta. Tig ihjäl och anmäl hatet, ägna istället din tid åt de tre följande punkterna.

2. Gilla och kommentera det som är bra

Engagera dig i innehållet du vill se mer av. Kommentera och gilla på innehåll som du gillar, då är det istället det innehållet som når fler. Ditt engagemang kommer också, som en bonus, att skapa glädje hos personen som skapat innehållet. En hälsning till din moster som köpt nya krukor är därför en väldigt bra investering. Dels blir hon glad, och dels har du injicerat lite trevlighet i plattformen.

Det har skapats en kultur av att man egentligen inte ska posta innehåll på Facebook om man är “cool”.Det finns flera aspekter kring varför det här är viktigt. En extremt kortsiktig effekt som vi vill åstadkomma är att rekommendationsalgoritmerna prioriterar det positiva och trevliga, gärna också det ofarliga, så att det visas för fler. En god cirkel istället för en ond. Men vi är mycket mer benägna att lyssna på personer i våra liv som är snälla mot oss, som bryr sig om oss och uppmärksammar det vi gör. Så om vi engagerar oss i våra medmänniskor kommer de dels att lyssna på oss men också göra mer som vi gör och närma sig våra värderingar för att de känner sig närmare oss.

Det har skapats en kultur av att man egentligen inte ska posta innehåll på Facebook om man är “cool”. Att man ska hålla bilderna på barnen för sig själv. Det är inte alls bra, eftersom det gör att vanliga vettiga människor syns i mycket mindre utsträckning än de egentligen utgör av Facebookpopulationen. Svälj coolheten och bjud på en snäll kommentar. Det gör att de som känner sig lite ensamma eller utsatta känner sig sedda av dig och de kommer vara mer positivt inställda till vad du säger i andra sammanhang. Vi vill alla känna oss uppskattade och viktiga, och allra helst del av ett socialt sammanhang.

3. Hjälp folk att hitta grupper de kommer gilla

Hjälp folk på Facebook att hitta in i grupper du tror att de är intresserade av och kan uppskatta (men som inte är extremt politiska). Vi har stora behov av att vara del av grupper och både familjen, kyrkan och idrottsrörelsen har varit en bra plats för folk att få detta behov tillgodosett. Men i och med att samhället förändras så förändras också folks grupptillhörigheter.

Många, som inte är Facebook-experter, hittar idag till grupper när de skapas i samband med nyhetshändelser som får stor spridning. Eller när de blir tillagda av en vän. Det gör att det är svårt för folk att hitta positiva och peppiga sammanhang som de själva gillar fastän de här sammanhangen egentligen finns. Om du aktivt hjälper folk du känner, och tror kan uppskatta en grupp, att hitta till dessa platser, är sannolikheten att tråkigare, trolliga grupper tar över deras vardag mycket mindre. Fråga alltid personerna i fråga så klart, och förklara varför du tänker på just dem. Det allra bästa är nog om du också är med i gruppen själv. När jag har studerat Facebook-grupper blir det också tydligt att många grupper med extrema åsikter är öppna, och därför lättare för folk att hitta till, än andra grupper som ofta är stängda eller hemliga.

Varje dag, vecka efter vecka, publiceras det flera inlägg som önskar de andra gruppmedlemmarna en trevlig dag.Det här kanske låter banalt, men det kan faktiskt göra större skillnad än du tror. Michela Del Vicario och hennes forskarkollegor har visat att vi på internet, precis som i verkliga livet, är mycket mer benägna att ta till oss information från människor som tillhör “samma grupp” som vi. När man tittar på extrema grupper av varierande slag kan man ofta se att de har gruppbeteenden som stärker gemenskap och tillhörighet för de som deltar. I flera av de stora grupperna med främlingsfientliga undertoner finns en tradition av att hälsa varandra godmorgon. Varje dag, vecka efter vecka, publiceras det flera inlägg som önskar de andra gruppmedlemmarna en trevlig dag. Om det här är den enda hälsningen du får på en dag blir det självklart en viktig händelse.

Sociala medier är en fantastisk plattform för att väldigt enkelt skapa tillhörighet och gemenskap. Så, om vi vill komma åt att många söker sig till främlingsfientliga eller trolliga grupper behöver vi helt enkelt hjälpa folk att tillgodose behovet av tillhörighet på något annat sätt.

4. Posta själv

Publicera eget innehåll som folk kan engagera sig i. De senaste åren har mängden innehåll som användare postar till sina flöden minskat drastiskt. Det gör att mycket av de innehåll som folk får upp i sina flöden spelar starkt på känslor eller har en agenda av något slag. Det är reklam, nyhetsartiklar och upprörda medmänniskor. Dela med dig av bra, vettiga artiklar du läser. Fråga dina vänner om restips, middagstips och barnbokstips.

Om varenda kommentar som idag läggs på att “prioritera upp” rasistiskt innehåll hos Facebook istället var början på en tråd om bästa receptet på saffransbullar, skulle räckvidden för det rasistiska innehållet gå ner drastiskt. Pinga in dina vänner, dela andras poster och skapa positivt innehåll på plattformen.

Slutligen

Jag förstår att det här känns helfel om man har suttit och debatterat med rasister dag ut och dag in i ett år. Men det gäller att förstå vilka spelregler som gäller på Facebook.

Och jag skulle nog ändå vilja säga att om det är dina egna Facebook-vänner som uttalar sig rasistiskt eller delar falska artiklar, så ska du alltid säga till och identifiera vad som är fel. Då har ni har en relation där du sannolikt har möjlighet att påverka deras tankebanor mer än någon som inte är deras Facebook-vän. Men prata inte bara med den här personen när hen säger rasistiska grejer eller delar falska artiklar. Prata med dem om allt annat också, så är sannolikheten att de lyssnar på dig mycket större när du väl påpekar att saker och ting inte riktigt ligger till som de tror.

Därför är Filterbubblan inte det vi borde prata om

Anna Loverus on November 24, 2017

Filterbubblor har blivit en populär förklaring till varför Trump blev president. Men få verkar veta vad en filterbubbla faktiskt är. Och istället för att debattera filterbubblans existens borde vi egentligen fokusera på något mycket viktigare.

När Eli Pariser 2011 definierade begreppet “filterbubbla” i sin bok The Filter Bubble: What The Internet Is Hiding From You, var det inte jättemånga som brydde sig om det. De senaste åren har begreppet dock blivit allt vanligare för att beskriva ett suboptimalt internet, särskilt i relation till demokratifrågor. Problemet är bara att många inte vet vad en filterbubbla är. Begreppet används ofta felaktigt för att beskriva generell polarisering på internet, det vill säga att människor med en viss politisk åsikt inte möter personer med en annan politisk åsikt, och kanske inte ens exponeras för media eller information som går emot deras befintliga världsbild.

Enligt många mätningar verkar polariseringen öka, både i samhället och på nätet, och World Economic Forum anser att polariseringen är ett av de största hoten mot mänskligheten just nu. Men den här trenden har förvånansvärt lite med eventuella filterbubblor att göra, polariseringen började långt innan Internet och sociala nätverk var del av vår vardag. Jag ska försöka förklara.

Eli Pariser, som har definierat begreppet, pratar om de tjänster som ger dig personliga rekommendationer av innehåll utifrån dina tidigare beteenden. Det här är något som började hos Google och andra sökmotorer. Varje individ som letar information på internet med hjälp av de här tjänsterna får se visst innehåll men inte annat. De hamnar i en bubbla där visst innehåll väljs bort för att dina träffar ska bli mer relevanta för dig, men du vet inte om vad som är bortvalt. Idag använder nästan alla sociala nätverk och sökmotorer en algoritm som räknar ut och visar vad som är relevant för dig, snarare än bara visar upp posterna i ordningen de publicerades.

Det finns flera missuppfattningar om filterbubblor som skapar stor begreppsförvirring, även hos de personer som debatterar ämnet. Därför tänkte jag räta ut några av de här missförstånden. Eftersom begreppet ofta väcker en intuitiv idé om vad det är så är det inte konstigt att folk hör det, och tror att de vet vad det är. Särskilt eftersom mycket av debatten handlar om huruvida det finns bubblor eller inte, känns det prioriterat att vi vet vad en bubbla är innan vi ger oss in i detta.

Alla har sin egen filterbubbla. Själva definitionen av filterbubblor utgår från att två personer som är tillsynes väldigt lika får olika sökresultat på samma sökord. Men du delar inte din filterbubbla med andra människor.
Filterbubblor handlar inte om politiska åsikter eller ideologier i sig, utan om all information. När det förekommer i sammanhang som rör politik eller samhällsfrågor på olika sätt skapar det andra effekter än om det handlar om effekterna av att dricka kaffe. Men det förekommer även där. Olika personer får olika saker beroende på vilka de är, men de vet inte varför. Politisk preferens är bara en del av all information om dig som en algoritm eventuellt använder för att ge dig relevanta rekommendationer, mycket styr dig sannolikt hårdare än politisk preferens, särskilt i Sverige där partierna är relativt många och lika.
Filterbubblor handlar inte om att människor självmant kategoriserar sig med likasinnade, exempelvis genom att aktivt följa sidor eller vara med i grupper på Facebook. Det är här det börjar bli svårt. För grupperna du är med i och sidorna du gillar påverkar vad du får upp för resultat när du exempelvis söker information, de är med och formar dina personliga rekommendationer, din filterbubbla. Men att du får upp massor av innehåll i dina flöden från en viss grupp som du är medlem i är helt naturligt, du vet varför det sker. Även om du får så mycket notiser från en viss grupp att din telefon tar time-out och allt annat innehåll trycks undan, så kan du alltid gå ur gruppen, och få annat innehåll. Men du kan inte gå ur din filterbubbla, för när du söker på fönsterputs, sockerkaka och hur man stavar “egentligen” och dina resultat ser annorlunda ut än mina, så förväntar du dig inte att dina resultat ska vara påverkade av huruvida du surfar från telefonen, eller av vilket postnummer du bor på. Filterbubblor handlar om att vi inte vet vilken information vi inte har, särskilt när vi tror att det vi ser är vad alla ser.

Det finns forskning som visar att polariseringen ökar när människor kategoriserar sig i grupper där alla tycker väldigt lika och att detta har ökat i och med internet. Men dessa grupper är inte i sig några filterbubblor.

Att kategorisera sig i grupper med likasinnade är något forskningen vetat länge att människan gillar att göra. Långt innan internet påbörjades ett skifte från tajta nätverk som var fokuserade kring familjen till en lösare och mer fragmenterad nätverksstruktur. Den här är något som förstärks av sociala nätverk, där det är lätt att hitta andra personer som ser världen som du, även om de inte bor i ditt kvarter.

Debatten om filterbubblor

I debatten om filterbubblor används ofta dagstidningen som exempel på att fenomenet inte är nytt eller att de inte finns. Du hade en socialistisk eller liberal morgontidning, kanske även en kvällstidning, och det var dina enda perspektiv. Men tidningsexemplet missar nästan alltid två aspekter när det används för att förklara att filterbubblor inte är något nytt:

Du visste vilket perspektiv din tidning förmedlade, det var även något tidningen själv kommunicerade.
Du kunde alltid gå och köpa ytterligare tidningar om du vill veta vilka perspektiv du inte fick.

Men dagens rekommendationsalgoritmer skiljer sig på båda dessa punkter:

Du har ingen aning om varför just du får just de träffar som dyker upp när du letar viss information.
Du kan inte ta reda på vilken information du hade fått om du varit någon annan.

De här två faktorerna i sig är varför vi började prata om filterbubblor över huvud taget.

När forskaren Peter Dahlgren står på Internetdagarnas konferensscen och pratar om att filterbubblor inte finns, argumenterar han genom att säga att vi har mer information att tillgå idag, än någonsin tidigare, och att vi kan mäta att folk får information från fler olika källor. Det stämmer. Men eftersom filterbubblor egentligen inte handlar om antalet perspektiv vi möts av blir det problematiskt när det måttet används som bevis för filterbubblor eventuella existens. Vi kan helt enkelt inte mäta huruvida filterbubblor finns genom att mäta antal nyhetssajter vi tar del av. Vi kan mäta huruvida filterbubblor finns genom att titta på vilken information som sorteras bort när folk försöker aktivt ta reda på hur något ligger till, och hur mycket två individers sökresultat varierar när de söker svar på samma fråga.

Filterbubblor handlar alltså inte om huruvida informationen vi tar del av är politiskt vinklad eller neutral. Utan om urvalet av information. Det är ett reellt problem som kan uppstå när tjänster använder algoritmer som ska rekommendera oss innehåll som vi sannolikt gillar. Det blir också viktigt att separera på passiv och aktiv konsumtion av information. För om vi aktivt letar information i ett visst ämne, exempelvis via Google, förväntar vi oss sannolikt ett större oberoende i våra sökresultaten, jämfört med när vi passivt konsumerar länkar som når oss i nyhetsflödet på Facebook.

Finns de då, Filterbubblorna?

En mycket snabb insamling av skärmdumpar på ett och samma sökord visar att det finns vissa skillnader. Men i mitt mycket bristande urval syns inte i närheten på så stora skillnader som tidigare visats upp vid jämförelser av Googlesökningar.

Mer intressant, när det kommer till sociala mediers påverkan på demokratin är hur folk själva kategoriserar sig i grupper med likasinnade. Det är en aspekt som aldrig kommer få lika stort medieutrymme eftersom det inte stämmer med narrativet att teknik är farlig eller antidemokratisk. Vi styr alla våra flöden i den riktning som passar vår individuella världsbild, helt utan hjälp från algoritmer.

Här är effekten på våra nyhetsflöden mycket större, beteendet är inte alls lika gömt i det dolda och vi vet att det bidrar starkt till polariseringen. Men frågan är – vad ska vi kalla det fenomenet?

Dark Posts in Social Media, what is it?

Anna Loverus on April 25, 2017

One buzzword currently circling marketing departments is “Dark Post”. It sounds like something dangerous and illegal, probably because you have heard about the “darknet” or the “dark web”. However, the two have nothing to do with each other.

Dark posts are nothing unlawful or dangerous; they show up daily in our Facebook feeds. Moreover, the only negative thing is that they are sometimes hard to track.

If you are active on Facebook now and then, you know that what you see isn’t only posts from other personal users. Pages – representing everything from your local flower shop, an extremist political party, or an initiative to teach kids to code – can also create posts that show up in your feed.

However, a lot of the posts that show up in your feed are ads. When a page creates a post on Facebook, it appears both on the page wall and in some followers feeds. The page can if it wants to, decide to boost the post to reach more people. Then they pay Facebook to show the post for Facebook users they want to reach. When this happens, the post technically becomes an ad*. These posts are often called boosted posts.

*If you do not boost your post it is referred to as an organic post.

You see dark posts all day long

Facebook advertisers want all their ads in your feed to look like any Facebook post. But the also want to send different messages to different users based on what they think they’ll like. To become relevant, or spend their money as effectively as possible, advertisers often create multiple versions of a message. They have different versions of copy, links and images, to find the one combination that performs best. This is a common advertising method called A/B testing.

But brands on Facebook usually don’t want their Facebook page wall to show the same post over and over with slightly different combinations of text, links and images. So creating an organic post and boosting it is not a very good option. Instead, they create a large variety of ad posts in Facebook Ads Manager and Facebook shows you the ad that it is most likely that you act on.

There’s a difference between a boosted post and an ad post created with Ads Manager. Boosted posts show up on the page wall, but ads don’t show up on the wall (if you don’t want them to). Dark posts are ad posts that don’t show up on a brands page wall. Dark posts live “undercover” or “in the dark” and no one except the ads targeted audience knows about the post. Another name for dark posts is “unpublished posts”.

You can create dark posts on both Facebook, Instagram, Pinterest, Snapchat and Twitter. On Snapchat and Instagram, all promoted posts are dark posts. The basic idea is the same to all these platforms, but this text will continue to talk about Facebook.

What a dark post looks like

On Facebook, boosted posts and post ads created in Ads manager look identical. You can see the “Sponsored”-mark directly under the page name when they show up in your feed.

You can only see a dark post if you are in the target audience for the particular ad and it happens to show up in your feed or if you know its direct URL. (You find the URL for a Facebook post by clicking on its timestamp).

Who sees a dark post?

Sponsored posts do not show up randomly. We all have different ads in our feeds, and the same ad does not show up for everyone. When a page creates an ad, they decide whom they want to show it to, and then it only shows up in their feeds. Targeting is the advertising term for choosing who to reach.

An ad shows up in your feed because you fall into a group the advertiser wants to reach. Either because of your gender, age, or where you live. However, it can also be because you have behaved a certain way online. Facebook use user data for targeting ads on its platform. Sometimes the pages you like or the pages your friends follow is motive for ad placement. But it is also possible to base targeting on different interests Facebook believe you have, most likely because you have clicked links or viewed videos about a subject.

However, even if the ads are showing up on Facebook they collect your online behaviours all around the internet. Facebook collects data from an enormous amount of sites online. Pages can re-target their web page visitors on Facebook if they have installed a Facebook script called a “Facebook Pixel”.

Dark post as a tool to reach niche audiences

Dark posts target particular niche audiences. The main idea is to create an ad and promote content towards someone likely to enjoy the content. The targeted user is often a potential customer, but dark posts can also be used to persuade voters in a presidential election or increase streams on a specific Netflix show among existing Netflix users.

For a brand publishing content on Facebook, it is sometimes hard to create content that talks to all your potential customers or fans at once. People are much more likely to engage with your brand if they feel like you are relevant to them, but how can you be of interest to a varied group of potential customers? This dilemma is why dark posts are often part of a successful content strategy.

Since an advertiser is in almost complete control over who sees an unpublished post, they can talk to multiple audiences at once, with different voices and propositions to each of them. The people who get the content in their feed do not know they are part of a bucket; they are just happy their Facebook ads are somewhat relevant**.

**Facebook also wants to show people relevant ads. They assign all ads on their platform a score from 1-10 based on their relevance to the target audience. If your ad shows a low relevance score, you can tweak your ad, or it’s targeting, to improve the effectiveness of the ad.

Dark posts in Social Media are not good or bad

Dark posts are in some circuits starting to become somewhat mythical, a tool used to manipulate people without their knowledge. The name for it is probably not helping though, especially not when it is few that know what it is.

The thing that makes dark posts somewhat criticised, or at least met with scepticism in the debate, is mainly two things: 1. Fake (junk) news sites, companies and political campaigns have used dark posts unethically in different ways, 2. It is very hard to track dark posts from an ad account that you do not own yourself.

First, the ethical aspects of dark posts

The fact that we can target people in extreme detail while being somewhat non-transparent makes it easy for an advertiser to enter into semi-legal or at least unethical activities either by mistake or knowingly. The regulations are often vague, and no custom exist around these issues.

Facebook is a profit-driven platform and seems happy as long as they get paid by advertisers. They have changed some policies related to fake news and ad posts after the storm of criticism that followed both the US presidential election and Brexit. With the new GDPR regulation from EU, arriving in 2018, this will probably change.

Second, transparency

Keeping track of what others are saying might seem like a vague argument against dark posts. Targeting ads have always been part of marketing and communication. But the vast amount of ads and the difficulty to track them makes it complicated. This grey zone is not a big deal when it comes to shoe sales, but if it is false political arguments getting spread around in the dark, it is somewhat problematic when they cannot be met or debunked.

It is also likely that we will see more tools for tracking dark posts popping up due to increasing demand from marketers to keep track of competitors. We can probably also expect greater transparency on Facebook, as well as on other digital platforms, both when it comes to both dark posts, and different types of online targeting. Facebook is trying to become an infrastructure more than just a social platform, but for that to happen, society needs essential insight into the platforms citizens use daily.

Should we be worried?

As with all tools, dark posts in the wrong hands can create some damage. However, it is not a dangerous tool in itself. If we keep spreading the knowledge about how dark posts work and how we can check if the messages we get online are really true, we don’t need to be afraid.

How we use Slack at home

Anna on November 15, 2015

Your setup of apps and tools for communication is as unique as your DNA. But when you’re in a relationship you spend most of your day communicating with only one person, so you might want to pick a tool together and make it yours. We decided to use Slack at home.

Since we are two geeks living together, we decided to create and move into our own private Slack-team. Slack is a communication tool for teams or organizations. So, when we decided to use it to communicate outside a typical work setting, it felt like a bit of a chance. A year later we’re happy with how it all turned out. And since we know that a lot of people are looking into doing the same transfer, we decided to share why and how we like it.

Why we like using Slack at home

Slack is multi-platform. When one of us decided to go for Android as the primary phone OS, we had to change from using iMessage. We realized we would change devices over time. And if we have to change how we communicate every time, we’ll get pretty tired.
Both of us already use Slack at work. First of all, Slack is very easy to use. My parents-in-law use it daily in their family group without any previous Slack experience. But it is already installed on all our devices, and that is off-course a bonus. And switching between a work context to our private team is simple and a practical feature.
We can have parallel chats, with each other, going on at once. In a text message conversation, it’s sometimes hard to find what you talked about just two hours ago. With Slack, we can organise our chats based on their topics. And everything is searchable if we’re looking to find something from a while back.
Slack handles links and other embeds neatly. A significant amount of what we share are articles and other interesting internet reads. Sometimes YouTube videos or music from Spotify. The fact that Slack display messages to highlight their content is a beautiful detail. But it’s practical when you need to find something or just to differentiate posts from each other. It might sound like a small feature but compared to just have the URLs this makes our chat-life way easier.
Slack saves and archives our conversations. Remember every time you lose/change your phone and suddenly are all your messages lost. Since your device is not related to the Slack messages you send, that won’t happen. And we will have everything we say safely stored for years to come.
Slack is free. The price was not a major argument for us since we only choose between free messenger tools. But others might care about this, so it goes on the list.

Our current Slack setup

You can set up your Slack environment pretty much as you like. There’s a lot of bots and integrations to look into if you’re missing a feature out of the box. This guy even integrated Slack with his online supermarket. Even though we know we’re able to do a lot of customizations if we want to our current setup is simple.

- Shopping. With Slack, we don’t need another tool for sharing shopping lists. We put in single items or create “snippets” with longer lists of what to buy. When you buy something, you mark it with an emoji. It works well both for everyday items and things we rarely buy.
- Photos. Ours and others, in one convenient place.
- Read/Listen/Watch. One channel for each type of content. It makes it easier to find what we’re looking for.
- Moving etc. When you’re moving in together with someone, you have plenty of discussions going on. Especially when you are selling two apartments and buying one. All ideas about new furniture and tedious electricity supplier research belong in this channel.
- Direct messages This is not a channel per se, but most of our communications are more of the type “How’s your day?” or “I’m buying pasta on my way home” so we use direct messaging a lot.

This setup is basic, but it works well for us. Sometimes simple makes it more practical. If you want to know more about the different bots and integrations you can use with Slack, take a look at their official collection. And if you want to know more about how we use Slack at home, drop a line in the comments or send me a tweet.