Personalisation works! Over the course of this post, you might be thinking I’m not a fan, but nothing could be further from the truth. It’s self-evident that customers respond well to an experience that is tailored to their specific needs and circumstances.

However…as we get further and further up the sales conversion funnel, focusing on people that aren’t yet customers, and trying to personalise their experience based on more limited information, things get a bit foggier. This post explores the different trade-offs being made, and why going all-in on personalisation might not lead to the happy ending everyone expects.

First, a couple of success stories. I’ve had direct experience of some fantastic personalisation initiatives that significantly moved the needle. For two of the airlines I’ve worked with, we changed the hero image on the home page based on past buying behaviour. For example, if you’d flown with kids you got a family shot. If you’d flown with a significant other, you got an image of a romantic dinner with a glamorous backdrop. Sales shot up.

And if you haven’t come across the Obama campaign A/B testing story, that’s an eye-opening quick read. It isn’t driven by personalisation, but serves to illustrate that tiny changes can make a big difference - it’s no wonder personalisation is such a seductive concept.

In the previous decade, much of the narrative in the Gartner and Forrester space focused around the importance of personalisation as a differentiating factor (although that’s swinging hard back the other way). This would inevitably lead to a platform selection process focused on which platform delivers the “best” personalisation capability (often an Adobe vs Sitecore bake-off, although other platforms joined the fray), and as part of that exploration, what are the licence costs?

But these licence costs are a drop in the ocean compared to the cost of building and running the system. And herein lies the rub.

For a previous client, we were implementing a program of personalisation that changed the web experience based on a cascading flow of what we knew. From the customer IP address, we could see where they lived, and what the weather was going to be like that weekend - and therefore what they might be interested in doing. We could then infer more from their search and browsing behaviour, scoring the user and matching them with products. And once they bought something, we knew even more about them, and could personalise the experience even more tightly.

But we hit a sticky patch along the way. For so many permutations of users, we needed to create a huge array of content, and this took attention away from other work. We needed to score every page on the site so it could be matched to the right customer. Was it all working though? Well, we needed an even more sophisticated data analysis system to understand the firehose of data. And analysts to interpret this data. And processes in place to review these insights and approve recommendations. And as the system became more complex, the questions that arose needed input from more people - the client needed to go through a restructure. All of which took focus away from the core focus of serving customers and turning a profit.

The system ground to a halt under its own complexity.

So where did it go awry? The step that gets missed out, time and time again, is answering the question: thinking holistically, what is the point of diminishing returns?

This is a moment of truth. It’s a great question to ask, because it forces you to think through the implications of your decisions and form an opinion about what you’ll get out of the venture and what you should therefore put in.

Anecdotally, I’ve heard of several medium-sized and large organisations reviewing their platform selections from 5-10 years ago and finding that they can no longer justify the inclusion of Sitecore or Adobe. This decision is driven by a few factors, but the inability to unlock the benefits has been central to this trend. As mentioned in a previous post, only about 40% of Sitecore customers have implemented personalisation, and my experience tells me that this is because the cost of change is so high.

It’s hard to get an independent perspective in this space. Adobe and Sitecore sell a very compelling dream, and the sessions at their conferences focus on how to implement rather than if. And most of the larger agencies are staffed with developers expert at implementing the larger platforms, and they need to be fed more of the same work - so you might not get an independent view there. And the larger consultancies will recommend the bigger platforms even though they don’t have first-hand experience of building them, never mind running them.

And none of them got rich by recommending you do less work.

BUT, to bring it back full circle, let me say it again - personalisation works. As long as the benefits and trade-offs are properly understood, it can play an important role in the growth of your business. So here are a few tips for how to structure your thinking and organisation to extract the value.

• If you’re thinking of moving from Sitecore or Adobe, take pause. They’re good tools if used in the right way, and if you jump to another platform without understanding the root cause of failure to extract value you might find yourself in the same situation further down the track. And it might simply be that the complexity needs to be scaled back.
• If you’re going through a platform selection process now, especially if you’re in a mid-market company, think through how you’ll operationalise the platform before you select it. You might reach the conclusion that you don’t need a heavyweight platform like Adobe or Sitecore, and instead you could manage just fine with a combination of an open source platform plus a lightweight personalisation tool (e.g. Umbraco and Optimizely). The limitations of these platforms could serve as a helpful way to keep complexity under control, and drive focus within the team - as well as unlocking budget to be spent elsewhere.
• You might even perform a quick exercise before settling on your platform; build a point of view on what you’ll want to personalise, prioritising the opportunities up front. At the very least you’ll discover how difficult it can be to come up with this plan, and you’ll also have a set of requirements that you can bake in during the build, rather than having to retrofit this capability later on. Remember; all personalisation is data-driven, and uncovering this data framework early in the process can help you avoid costly mistakes.
• Find a way to help your teams stay focused on your customer and your strategic point of difference. By staying true to this North star, you’re more likely to resist the siren call of deeper and deeper personalisation with ever-diminishing returns.
• Consider the broader digital ecosystem as well. It’s vital to get your digital strategy and roadmap firmed up with this bigger picture in mind, before making platform choices. For example, if you’re considering providing a customer experience / journey across multiple touchpoints (voice, chatbots, etc.), then you might choose to drive everything from APIs, which in turn will influence all your other decisions. And the emerging trends around Static Site Generators / the JAMStack, and the CMS’s aligned with this, afford new opportunities to improve performance, security and scalability.
  
  

If you’d like an independent perspective, we’re happy to shoot the breeze. Let us know if you’d like a chat.