But pretty soon descend into chaos…
Boutique Boutique Boutique Casual Casual Skirt Skirt Casual
These interests come from all sorts of interactions on Facebook including links you’ve clicked on or videos you’ve watched so you’re likely to find Facebook thinks you’re interested in plenty of things that you wouldn’t exactly list on your CV. As a result targeting people by a single interest is prone to reaching the wrong people.
So Facebook thinks I’m interested in digital marketing but that could be because I clicked on an article a friend sent me 4 years ago that I spent 2 seconds looking at. The same could go for Adwords or Entrpeneurship, so instead of using an audience like this where I might only match 1 of your targeted interests:
I would use this audience to target people like myself:
Casual Skirt Boutique Casual Boutique Casual Boutique Skirt Boutique Casual Skirt Skirt Boutique Boutique Casual Casual
But what I’ve found about these interests is that although some may be way out of left field, others are bang on. So if you’re advertising to someone with one relevant interest it’s highly likely to be wrong. If you’re advertising to someone who shares 3 or 4 relevant interests it’s highly likely to be accurate.
Casual Casual Boutique Casual Skirt Boutique Boutique Skirt Obviously to make this approach work you need to invest a bit of time playing with interests and finding a combination that gives you a decent sized pool but remember you can use multiple interests at each layer of the stack and those will be matched as OR rather than and. So if I find each of my interests are a bit too narrow I can expand them at each level like you see below:
Strategy 4: Stacking Interests With Brands
Really similar to the above but instead of using “interests” per se I’m going to use popular brand names, which also show up as interests in Facebook Ad targeting, to narrow down my audience.
Using brands as interests in a stack is really powerful because it again limits the margin for error in Facebook’s classification of audiences.
So lets say for example I want to target small business owners but I’m not sure how accurate Facebook’s data is on the broad “small business” interest category. I can refine that targeting by only going after people who also have an interest in brands that I’d associate with small business, in this case accountancy packages…
Here’s another example, the “physical fitness” category is massive and very broad, probably not that useful for an advertiser. But combine it with some brands that suggest the person is serious about fitness and you massively increase the chances of Facebook being able to reach a real fitness enthusiast.
Strategy 5: Using Interests as Demographic Indicators
If you’ve played around with Facebooks demographic bands you’ve probably like me had mixed results. In the UK we have Acxiom data in Facebook Ads which lets us target homeowners, high net worths and so forth. But I don’t trust that data entirely.
The way old school advertisers targeted demographics was through their interests and that’s still my preferred approach. The things people “like” and particularly the brands they follow and the publications they read are massive indicators of their demographic bandings, social class, lifestyle etc so stacking these types of interests on top of other types of targeting is massively powerful.
So for example you would see vastly different results targeting this audience who may be interested in fashion & are also interested in sites like Buzzfeed:
Than you would with the same interest but stacked with the Wall Street Journal:
We’ve seen really interesting results stacking things like publications and supermarket brands on top of broad interests in this way.
Strategy 6: Build Audiences of Sharers For Viral Campaigns
Strictly speaking this strategy is more likely to use demographics data rather than interests but there’s cross over.
Facebook Ads is a great platform for launching content that you want to get in front of sharers and influencers. Take a look at the audience below from a real campaign we’re running to boost sharing. Data on bloggers, vloggers, journalists and other types of writer and internet folk is really accurate because unlike the pure interests we’ve been using in the first 5 strategies these are more likely to be self entered by the user.
We find this audience have a much much higher propensity to like and share content than ‘normal’ facebook users and they tend to have bigger and more active networks, so their reactions to your posts are going to spread further organically. Try it out when you’re next promoting a post that you want to ‘go viral’.
Time to Get Back Into Interest Targeting
So if you’re using interest targeting already, or if you’ve tried it and not got the results you want time to think again. To recap what you should do now:
- Start splitting down large audiences based on interests and serve those people ads they’re more likely to be interested in clicking on
- Start stacking interests on top of each other to remove Facebook’s margin or error
- And refining your targeting by adding lifestyle brands or publications as modifiers to reach more of the right sort of people.