One offer, 5x the revenue: AI-powered Next Best Offer

Our solution? An AI-powered Next Best Offer (NBO) engine that can anticipate what each shopper is likely to buy next, whether they are browsing the site, opening an email, or walking into a store. The system didn’t just improve targeting by combining real-time behavioural data with deep learning. It redefined it.

The results were swift and dramatic: a 500% increase in revenue from NBO-driven communications compared to standard promotions. Customers clicked more, converted faster, and came back more often, all because the offers felt less like marketing and more like relevance.

This is just a taste of what your brand can achieve when you shift your personalisation strategy from broad strokes to behavioural precision. It’s no longer enough to rely on guesswork or gut instinct when it comes to personalisation. But the truth is, you likely have what you need to get started with AI-driven, personalised offers: data your customers are already giving you.

A Next Best Offer (NBO) is exactly what it sounds like: the smartest, most relevant thing you can put in front of a customer right now. It could be a product, a promo, a piece of content, or a well-timed reminder. Basically, it’s whatever the AI model predicts will drive the highest likelihood of engagement based on everything that the customer has done before.

It’s not just a souped-up recommendation engine. NBO is built to replicate what a great store associate does: understand the customer, anticipate their needs, and make timely, helpful suggestions that feel personal but not pushy.

Here’s how our NBO tool works behind the scenes:

Instead of grouping people into broad personas or relying on rules like “customers who bought X also like Y,” NBO uses deep learning to recognise individual patterns. What products does someone gravitate toward? How often do they buy? What channel do they convert on?

The model digests all of this data and more, then scores each possible offer in real time. So, rather than recommending a beginner skincare bundle to someone who has shown interest in the “beauty” segment, you’re highlighting the serum they’re most likely to buy, right when they’re most likely to need it, on their personalised app homepage. 

NBO isn’t just about what you show—it’s when. If a customer typically replenishes every 90 days, the system won’t wait for them to come back. It’ll proactively tee up an offer via their preferred channel—maybe an email, maybe a push notification, maybe something in-store. The goal is to meet them just before they realise they’re ready to buy.

Web. Email. App. In-store POS. A well-integrated NBO engine doesn’t treat these as silos—it sees them as signals and surfaces. If a customer browses online, doesn’t convert, and then opens an email later that day, the offer should reflect that entire journey. The experience should feel coordinated, not fragmented.

This kind of orchestration is where most brands fall short. You might have great product recommendations on your homepage, but your email still blasts a generic sale. Integrating an NBO solution into your martech stack fixes that disconnect by aligning your message across every channel, driven by what the customer actually needs next, not just what has too much inventory.

For the cosmetics brand we mentioned earlier, the personalisation challenge wasn’t a lack of data, but a lack of coordination. They had loyal customers, a robust omnichannel presence, and strong creative. But their marketing still relied on generalised campaigns and static rules. Our goal was to bring precision to that chaos and make every customer interaction feel personal, contextual, and timely.

We started by designing a custom deep learning model trained on real customer behavior, including online browsing history, in-store purchases, mobile app activity, email engagement, and loyalty data. Instead of treating every shopper the same, the model learned each customer’s rhythm—what they bought, when they bought it, and how they liked to engage.

In action, this looked like a customer who typically buys skincare every six weeks and opens push notifications. They’d get a personalised app alert featuring the next product in their routine, just before their usual restock date. Or a fragrance buyer who likes to shop for minis during gift-giving holidays might open their inbox to a timely offer with seasonal picks they hadn’t tried yet.

What made this work wasn’t just the intelligence—it was the integration. The predictions weren’t stuck in a dashboard or buried in a business intelligence tool. They were piped directly into the e-commerce site, email system, loyalty app, and even in-store clienteling tools. No matter how a customer interacted with the brand, they were met with the most relevant offer, in the most natural place.

The outcome was a 500% increase in revenue from NBO-driven communications compared to mass promotions. But even more important, the brand saw a meaningful lift in loyalty and engagement. Customers felt seen and supported rather than simply sold to. And because the system continuously learns and adapts, its performance only improves over time.

You don’t need perfect data or an enterprise AI setup to start reaping the benefits of a Next Best Offer solution. One of the biggest misconceptions about personalisation at this level is that it requires an all-or-nothing investment. In reality, the most effective NBO strategies often begin as focused pilots with clear, measurable goals.

Start by taking stock of the data you already have. You likely have more than enough, especially if you can access purchase history, email engagement, and basic site behavior tied to individual customer IDs. What matters most at this stage isn’t completeness, but usability. Clean, consistent inputs beat massive but messy datasets every time.

From there, identify one high-impact use case. Maybe it’s swapping out the default product block in your email with a personalised recommendation. Or updating your homepage to show each returning visitor something they’re statistically more likely to click on. By keeping the scope tight, you can move quickly and see results without overcomplicating the setup.

Then it’s time to run a controlled test. Compare performance between the NBO-driven experience and your business-as-usual approach. Are people clicking, buying more, or engaging more often? This kind of real-world validation is what builds the case to invest further and refine your model over time.

And no, you don’t have to build it all from scratch. Whether you’re working with DEPT® or leveraging existing tools from your cloud provider or other personalisation platforms, you’ll find different ways to plug NBO into your existing stack. What’s most important is getting into the market and starting to gather learnings from real behaviour so you can iterate and optimise. 

With NBO, every test makes the model smarter, each data point sharpens the predictions, and all the small steps get you closer to a truly personalised experience.

AI-powered Next Best Offer recommendations aren’t just a tactic. They’re a smarter way to build relationships, increase revenue, and future-proof your marketing engine. As consumers grow more selective about the brands they engage with, relevance becomes your most valuable currency. And relevance doesn’t come from louder messaging. It comes from better timing, better context, and better understanding of what your customer actually wants.

The brands leading the way aren’t waiting for the perfect setup. They’re experimenting, learning, and putting data to work to unlock better experiences. And they’re letting AI do what it does best: spot the patterns, predict the moments, and guide the next move.

If you’ve got the data, you’ve got the foundation. With the right AI transformation strategy and a clear use case, you can build an engine that not only performs—but learns and improves with every interaction.