Content Remarketing: an experiment with Resonance, Part 1
As we whinged about in our Crap slideshare, it’s getting harder and harder to get content discovered and downloaded. That means content marketers will have to get better and better at the kinds of things we thought we’d left behind: like dreaded outbound, interruption-based advertising (gasp!).
That actually doesn’t bother me: if we do it intelligently, advertising that offers content will out-perform product-based advertising and will still be serving the prospect, in the spirit of content marketing and inbound and all things holy.
So when Marcin Grodzicki set up a call to discuss a new content retargeting platform his team was building (now called Resonance), I was happy to take the call. And by the end of the call, I was eager to give his new platform a try.
Definition break: “Retargeting marks or tags online users who visit a certain brand website with a pixel or a cookie and then serves banner ads only to the people who have shown at least some amount of engagement in the original brand.”
I know, I know: if advertising is a dirty word these days then, to many people, ad retargeting is like animal porn: unthinkably yukky.
I’m a card-carrying hater of crass retargeting (as posted here, “When retargeting turns into stalking’). But I actually believe that retargeting (or remarketing if you prefer) is a powerful technique that can really improve some of the metrics content marketers care most about (like traffic, engagement and shares). As long as it’s done intelligently.
Why I think B2B remarketing has a role to play in content marketing
- Because something like 90% of your web visitors will never make themselves known to you – and may never come back.
- Because people who make it to your website even once are worth the effort – they’ve already responded to something you’ve done. That makes them much more likely to be prospects than people who haven’t.
- Because the data speaks for itself – our friends at Mediaplex tell us that, in B2C, the numbers are pretty indisputable: done right, retargeting can dramatically out-perform normal advertising.
So combining retargeting with content marketing – chasing folks who have visited your site with content offers – should be a pretty powerful combination.
The downsides of B2B content retargeting
- It can be creepy. It’s weird to have a pair of shoes I looked at on a brand’s website follow me around the web for days or even weeks.
- You have to pay for it. So it’s competing with every other way to buy eyeballs and fingertips.
- You risk preaching to the converted – some of the people you retarget have already bought (or read your content). So it can look like you’re wasting money chasing them (even if you are only paying for clicks, not impressions). For some brands (like awesome B2B content marketing agencies) this perception itself can be damaging. It makes you look like an amateur.
So what I want to find out is this: how big is the upside of content retargeting? How big is the downside? And are there ways that content marketers can harvest the upside while minimising the downside?
The Resonance approach
Resonance’s big idea is that content marketers have ‘content funnels’ – and that a prime goal should be to get people to consume as much of our content a possible, moving down the funnel as they go. So if we can get more people to consume all of the core content pieces in our funnel, we’ll be turning more casual visitors into prospects.
With normal retargeting, you capture all users of the site and retarget them with one or many ad units. The trouble is, you have little control over who sees what so people will be exposed to the same content constantly and will soon develop banner blindness (or get really annoyed).
So the content funnel approach is a very cool spin on standard content retargeting: understand what content each person has already seen so you can offer them the next most relevant piece (instead of chasing everyone with the same ads). But does it work?
Let’s find out: the experiment.
We need to know the answers to the above questions before we can recommend this new kind of content remarketing to our clients. So we asked Marcin if he could run an experiment, using the Resonance platform to run a content remarketing program for our own content library. He agreed, so we put aside a micro-budget to run a test: looking at retargeting versus normal (un-retargeted) ad banners; and at content versus product ads. So people that have seen the Content Marketing Manifesto will be retargeted with ads for The Big Fat Content Marketing Strategy Checklist and so on, until they have seen (clicked through to) all of our content.
A caveat: our traffic numbers aren’t exactly Amazonian and our budget was even smaller, so it can be hard to get solid numbers. But we hope to at least get an indication as to whether or not content retargeting works.
Phase 1: Product Ad Retargeting
We created an ordinary retargeting campaign for two of our most important services: B2B Strategy and B2B Content Marketing. The campaign setup cookied our site visitors and put them on a retargeting list. Then we showed those people banners for our products across display advertising inventory all over the web (including sites like Wired, CNet, TheDrum, VentureBeat, Mashable or Youtube). With the help of the Resonance team, we tracked all of the key metrics of the campaign, like price of impressions or click-through rates. We also analysed how the traffic we generated performed on our site. (We’ll report back on this in a future post).
Now it’s time for the sinister Phase 2 of the experiment:
Phase 2: Content Ad Retargeting
In the second phase we’re taking people on the content journey – guiding them through our eBooks and Slideshares and making sure that each visitor gets the most relevant content exposure. Banner creative will be customised for each content piece and served in an ‘advertorial’ style. (We will also sample our users and see if we can select the ‘best performing’ group based on the way they interact with our content).
Our focus will be on improving all four metrics: Pages per visit, Time on site, Bounce rate and Conversion.
So far, the results are really interesting. But this phase has just started, so we’ll report back in the second post on this experiment.
A few bits on the ‘How’ part
We left the ads themselves to Resonance because we wanted to take creative out of the equation: they just cranked out some simple ad banners and then used multi-variate testing to discover the best-performing ad units.
We then placed two retargeting tags on the content pages selected – using Google Analytics to decide on the best candidates
And dropped some analytics code in the footer of our site – so we can track our returning visitors across every page
We couldn’t select what sites our ads appeared on – this will be possible but it isn’t yet. That means we popped up in some very odd places.
One worry is that popping up on sites that have no relevance to content marketing makes us look like bad marketers. In truth, it’s the visitor that dragged the ad behind them, we didn’t buy inventory on the irrelevant site. But it looks like we did and that feels weird. It also jumps off the page and sings, “Retargeting happening here!” in a way that doesn’t happen when the ad is served on a relevant site.
Creatively, we might have some fun with this, doing self-referential ads that say things like, “It’s us again!” — remember, our target prospects are marketers, so they’ll be in on the joke (most audiences won’t).
So these ads on Mashable:
Look a lot more natural than this ad on YouTube:
Or this one, that appears to use the Queen as a Velocity spokesperson:
Or this one, that uses Richard Branson to shill Velocity:
An unsolicited Skype conversation with a client who got cookied:
Client: BTW – your digital remarketing efforts are working nicely
Neil: It’s a test. Might recommend for the content hub… Where did you see it?
Client: I’ve had ad placements in some funky places over the last couple of days. There’s one in front of me on Accountancy Age.
Neil: Think it’s too much? Would be interested in your feedback.
Client: Not at the moment I don’t, at all
Neil: Cool. I’ll tell Doug, he’ll be chuffed. We’re thinking it could be good for clients overall