Today’s business landscape is challenging and complex. So with the promise that ‘data’ is the answer, and swayed by the promise of ROI that data can bring, CFOs are increasingly signing off on moving marketing budgets over to consultancies like Deloitte and Accenture.
Consultancies like this understand how to install big enterprise-grade software and thus data, and typically their ‘client’ is naturally the IT Department.
That means they can sometimes take a too-clinical approach and actually alienate customers.
By way of example: Brand X thinks its cool to send a mailer telling someone how much of their personal data it has been able to collect and that it’s offering its products based on what they know about the consumer.
That’s going to leave the consumer wondering how that brand knows so much about them. It’s creepy and invasive and now they don’t want anything to do with Brand X.
And that’s Brand Fail.
It’s because of this kind of advertising that 75% of consumers find personalisation creepy, with 22% saying they’d take their business elsewhere.
So what’s the halfway house between knowledge-is-power and understanding how to deploy what you know about consumers in a way that truly does engage and lead to sales?
The real power of data-driven advertising comes when it’s helpful. Done correctly, it can have a ‘wow that was good timing’ sort of effect.
It should look serendipitous rather than forced.
Let’s say, for example, Thandi’s been looking for a BMW 3 series online, it would be far more helpful to see ads for relevant vehicle financing, with 1) an appealing offer, and 2) the exact car she is looking for dynamically (automatically, programmatically) appearing in the ad.
“Finance the car of your dreams, BMW 3 series at X% off”. Helpful, personalised, feels like great timing. “Wow, amazing timing!” says Thandi.
Even here though, it’s important not to get carried away with personalisation. Just because you’ve got a lot of data on someone, doesn’t mean you always have to use it. You need to use it well, construct your audiences and segments, and drive the right message at the right time.
But even if you’re careful about how much data you use, it can be easy to veer into creepiness. If someone feels like they’re the focus of the ad, rather than the product or service you’re trying to sell, they’re likely shut down.
The human touch
Perhaps most important, however, is to remember that the people viewing these ads are just that: people.
Let’s say you go into a business meeting with someone. Naturally, you would’ve done a reasonable amount of background research on them. But you wouldn’t reveal all that information in the first five minutes. That would make you look like a stalker.
Instead, you’d use what you know to gradually get the person you’re meeting to open up to you.
The same is true of the internet. If you’re an advertiser, people will become more and more comfortable with the interactions you have with them over time. Rush it, and you risk alienating them for good.
While it’s understandably tempting to chase sales as hard as possible, doing so makes it all too easy to forget that you’re targeting people.
Make it scientific
But how do you get the balance right? After all, it’s easy to be empathetic towards an individual potential customer. Doing so with thousands, or hundreds of thousands, of customers is much more difficult right?
It used to be, but not any more.
The answer here is to test extensively. By looking at which degrees of personalised advertising are best received, you can figure out what people are most comfortable with.
Testing can also tell you whether you’re advertising too frequently and whether customers are seeing your ads in too many places across the web.
Ultimately, it’s about realising that data isn’t just about finding out everything you can about your customers. It’s about using it to be helpful and to tweak your marketing strategy according to the feedback you get.
Numbers can tell you a lot about someone, but they can’t tell you everything. Unlocking curiosity still requires a human touch. Combine the two and you have a seriously powerful mix.
This article was first published by Memeburn on the 27 November 2018.