Enter a search term such as “mobile analytics” or browse our content using the filters above.
That’s not only a poor Scrabble score but we also couldn’t find any results matching
Check your spelling or try broadening your search.
Sorry about this, there is a problem with our search at the moment.
Please try again later.
Now that Twitter has rolled out functionality which is a lot like Facebook's Lookalike Audiences, it's time to take a closer look at what lookalike audiences are, how to use them and why you would.
Previously I've written about Facebook Custom Audiences, which create an advertising audience out of an email list. And I've also included 'lookalike audiences' as one of the key features that makes Facebook such a compelling place to advertise.
So far, though, I have avoided going into too much detail about lookalikes as they were somewhat unique to Facebook and required specialist knowledge.
But with Twitter's recent announcement that they, too, will support lookalike audiences, it's time to cover them in greater detail.
That is, there is now some agreement among the social media powerhouses that lookalike audiences are an effective strategy for reaching people on social media who may be interested in your product or service - so they are certainly worth learning more about.
Like with any new marketing strategy, though, they are not without controversy - and you will see why very quickly.
And though there are sensible reservations about using them, first read an overview of lookalike audiences through which you can learn a bit more about how they work and whether they might be useful for you.
What are lookalike audiences
For those who were previously unaware of lookalike audiences, here's a quick definition:
A lookalike audience is an algorithmically-assembled group of social network members who resemble, in some way, another group of members.
Things to note:
- The original, 'seed' group is a user list provided by you. This is not interest or behavior targeting.
- The seed group are identified by email address, phone number, website cookie, or device ID.
- You don't need to provide any other information about your seed group - just their identifier.
- You are not told by the social network who in your seed group has an account on the platform, but you are told how many do.
- The social network runs the algorithm and decides who of their members 'looks like' your seed group using what they know about your seed group members.
- You do not know why the lookalike group matches your original group. The algorithm is proprietary and secret.
- You are not given access to the names on the list, but you can advertise to them as a group (which is the point).
- Once created, they become an advertising audience like any other demographic, only they are unique to you because you created them.
As you read these notes, alarms bells may be going off.
Building a lookalike audience requires that you upload your most valuable asset (your customer or user list) to a social network. In return, you may, possibly, reach more like them. Yes, it sounds a bit crazy. And no, you shouldn't do such a thing lightly.
But for the length of this post, please put your suspicions and data privacy questions aside so that you can learn about how they work - and why they may, indeed, be the next big thing in digital marketing.
How to use them
There are plenty of step-by-step guides to building lookalikes on Facebook - including one from Facebook. And because the other guides are done so well, I won't repeat step-by-step instructions here.
There are, however, a few best practices to follow when creating them:
- The seed group must have at least 100 members.
- Quite often, though, the emails (or other IDs) that you have do not match members on the network. So you will probably have to upload more than 100 to be sure you can build a seed audience.
- You can optimize for greater similarity or for reach. It's sensible to create multiple lookalike audiences each with different levels of similarity.
- Note that it takes time for Facebook to run the algorithm to create your audience - but that time has been vastly reduced from overnight to typically < 1 hour.
- Make sure the members of your seed audience are as similar as possible. If you know age, product interests, location be sure to pre-segment your audience using those factors.
Now, when you use your final audience, you can only target an ad to one country at a time (but you can of course create multiple campaigns to hit multiple countries.) You can also filter your lookalike audience further using Facebook's interest, education, age, etc. ad filters.
Now this is a crucial point which you shouldn't miss. Lookalikes offer you a way of assembling a social media population who is likely to use your product or service - and then segmenting them further so that your ads are more relevant for them.
It's an incredibly powerful combination of targeting and segmenting that, really, has never been available before.
Why to use them
Now that you understand what lookalike audiences are and how they are implemented, I bet you already have some idea about why - and where - you might use them.
In case you don't, however, one reason is that there is something about your intended target audience which is not captured well in social media - say a 2nd language or level of disposable income. Luckily, though, you have an email list of people with those attributes. You can then upload them, create a lookalike audience and now you can do what the social network - and possibly your competitors - cannot.
A lookalike audience, then, offers you some hope to reach those valuable consumers so you can offer much more relevant ads and, hopefully, get a much higher quality response for your ad spend.
And I'm sure there are many other reasons to use them - but like with any new, previously unavailable marketing medium you should think creatively about how you can benefit from it and then run thorough tests.
Back to data privacy
I can't tell you whether using your list for lookalikes violates standard terms & conditions and data privacy laws, but what I do know is that marketers are always looking for ways to improve performance. And should lookalike audiences outperform interest, demographic, and location targeting, you can be sure that it will be adopted in your industry.
It may require a change in t's and c's - or even a total review of your customer acquisition strategy - but if you do performance marketing it will likely be worth the effort.
Hopefully this brief guide has been sufficient to pique your interest in lookalike audiences so that you can think about how you might use them for your own industry. I'd love to hear what other people think of their potential - or their danger - in the comments.