Using customer segmentation for campaigns can be like playing roulette

…you could be gambling away your brand’s customer experience and a lot of revenue too.

Arggh. I’ve received another promotional email that’s of exactly zero interest to me! Yes, I’m male between 40 and 50. Yes, I live in a city and own my apartment. Yes, I bought a laptop from you two years ago. But I would never buy a huge screen TV to hang on my wall; I hate huge screens.

A marketer and Data Scientist myself, I understand why I receive such emails: like many other companies, this retailer has used its customer segmentation to target its campaign.

I know a segment is defined by the profile of its average customer. But I’m not a true reflection of the average customer within that retailer’s chosen segment; I guess the segment I belong to is something like male below 50, city dweller, income above average and already bought Hi-Tech products. The thing is, I’m not interested in TVs and I’m very happy with my current phone. Yet they will continue to send me their offers none-the-less. (Actually, I won’t receive their messages anymore. I’ve just unsubscribed!)

My fellow marketers – we’re the ones running email campaigns. Should we (and can we) do better?

The era of automated self-service campaign targeting and planning

I believe we can do better. Instead of using pre-defined segments, the latest in AI technology for marketing can find the optimal targets for each and every campaign.

Don’t get me wrong: segmentation is great. It’s very useful when you want to put your whole customer base into nine or ten sub-groups to better understand who your high-level personas are. You can even go further and create micro segments, let’s say a few hundred. But how do you deal with your micro segmentation when you have tens, maybe hundreds of thousands of products and want to promote an item to just a (hopefully interested) subset of your customer base? (Let’s not take a step back with the average profile or score per segment).

Let’s put segmentation aside for a moment, and talk about a better way of targeting; a way that’s simple.

Each and every time you want to sell a product or a group of products, just create the optimal target for your campaign…

‘Sure…yeah, in theory’, you might say. ‘But how do we do that?’

Well, what if you could target not by segments of pre-defined people, but target based on deep intelligence informing you on who is likely to buy what, and when?

What if you could do even better? What if for each campaign, you could list the customers who are highly likely to buy the product you want to sell in the days following your campaign? That would be perfect – a marketer’s Nirvana – wouldn’t it?

Well, it’s no longer a theory. The reality is that brands like Thomas Cook, Lacoste and many others are optimizing their marketing campaigns today, using Deep AI-based marketing techniques!

A real life example

Say you’re a ‘generic’ retailer and want to sell video games. You have a new football game launching in the summer and want to run sales campaigns. If you use segmentation techniques and have a segment for gamers (based say on earlier game purchases or browsing), you’ll probably use that for targeting.

But it’s unlikely your ‘gamers’ segment will be so detailed as to contain only gamers whose behavior suggests an interest in that specific football game. Your segmented campaigns will also be targeting Mario Kart players, Warcraft players, Assassin’s creed players and other gamers. But a ‘gamer’ isn’t necessarily interested in all games. Sure, some will have an interest in your game, but many won’t. And in targeting those people who aren’t interested, with every email you chip away a little bit at their experience of your brand, risking opt-outs. And not only that, but there’ll be many other customers you didn’t target who would have bought. Shame you missed them…

Using Deep AI techniques, such as Splio, you would create your marketing campaign in a different, more efficient way. You wouldn’t use a ‘video game’ segment or a ‘propensity to buy video games’ score (if you had such a thing). You’d use Splio to select the product you want to sell (the football game) and in minutes Splio would list the customers who are very likely to buy it within days of your campaign.

So for your summer campaign (product launch time), Splio will find the best customers to target at that time. Most of them would probably be hard-core football gamers with a tendency to buy games as soon as they are released. But when you’re planning your Christmas campaign, Splio’s Deep AI algorithms will locate Christmas buyers, such as parents and grandparents who are shopping not for themselves, but for their children and grandchildren. You could also use Splio to locate customers who tend to buy a few months after launch (perhaps once the reviews are published).

Think differently …start today

Based on my 20 years’ experience in campaign management, I truly believe this is the Holy Grail we’ve been looking for: a simple, seamless, highly efficient way to find the future buyers for each campaign, no matter the product we want to promote. It’s time to stop bluntly sending campaigns to entire segments. From a business revenue perspective, sales via segmentation will always be sub-optimal and from a customer experience point of view, it’s a disaster.

The big question is does it work?

Below is a sample of the benefits some of our customers have experienced using Splio.

At Splio, my colleagues and I monitor the impact our solution is having on our 70+ clients. On average, across hundreds of A/B test campaigns, our clients won a 49% increase in campaign revenues. We’re pretty proud of that!

I think segmentation is great for strategic marketing planning and customer insight. But if you want to sell products, you should use a solution that simply does that. Deep AI finds the people who are going to buy your product in the next few days. Seems simple? It is. Contact us for more details.