DIY on the e-commerce plumbing - how B&Q is honing its online customer journey to prevent buyers leaking out

If you've ever visited a B&Q DIY outlet, you'll know the feeling of wandering the aisles, looking up at massive shelves towering skywards, in search of the product you're after. Now translate that to digital!

B&Q is the largest UK home improvement companies with 300+ stores nationwide and over a million products sold online. It's part of the Kingfisher Group, Europe's largest home retail group, with 1,300 stores across eight countries.

It’s a massive operation, but that’s all part of its appeal, according to Ali Saker, Ecommerce, Customer Journey & Strategy Manager at B&Q:

A wonderful thing about working at Kingfisher is that scale is something that we have in abundance. We're always working with amazing people that are cross-banners and we're all learning different things all the time, which is brilliant.

But size brings its own issues to overcome, she adds:

The challenge with that scale is that you can often find that really useful information is in a silo. So one of the things that is great...is actually finding those opportunities for us to collaborate. We recently ran a cross-banner training session, where we we came together to do a bit refresher training, but what we also did was start talking about, 'What have we well, what have we been doing recently? What insights have we found? How's that applicable across our banners?'.

Having that opportunity to share and talk and collaborate helps when it comes to common agendas across different parts of the Group, she suggests:

Democratizing data is something that we're really aiming at looking at internally [at B&Q] as well as cross-banner, so we can start partnering with other analytics teams, insight teams and research teams, so that we're really efficient, [and] we’re not all researching the same thing. We might be building on what somebody else has already looked at.

Journeys

In her own role at B&Q, Sarker has spent a lot of time working on what the firm terms ‘vertical journeys’ by customers. She explains:

We have loads of different categories that we look after. What we do is look at [things] category by category. So we might take kitchens, for example, and we'll take it out and just go, 'What is the end-to-end of this?’. We'll start with quant [quantity] first - where's the traffic coming from, where does it land, where does it go, what's happening when customers land, and where do they go?

To this quantitative approach is then added a qualitative layer for context:

You get the KPIs you're hoping for out the other end, and that will give you a picture, one picture, that might give you some drop off points. It doesn't give you the full richness. So what we do then is we layer on top some qualitative data. We'll have a conversation with people who have either just built a new kitchen, or they're in the middle of a kitchen project, and we'll say, 'What is it that drove you to start a kitchen project? What inspired you? What questions did you have in your head at the time, and what are you expecting from a retailer?'.

Then we'll ask them to have a little look at our website and our apps, and we'll ask them to look at our competitors and say, 'Well, are we, as a industry. answering your questions, or are we creating a few more?]. It's only really once you've done that that you get the full idea of what that end-to-end journey is.

With that information to hand, consideration of what constitutes best-in-class for customers takes shape, Sarker says:

We'll think, 'OK, if you weren't going to do this online and you weren't going to touch a store, what would be the best we could possibly be?'. And then we go, 'We've got 300 stores. We're B&Q, we've got loads of USPs. How do we layer that on?’, and we create what we call like the step change for that category. But until you've done the quant and the qual, you don't know what your metrics going to be. You don't know what your customer problem is, so you've got to do that piece first.

Plumbing for success

A good example of a vertical journey would be B&Q’s plumbing offerings. Sarker says:

The interesting thing with plumbing is that we didn't necessarily start because we had any insight. What we had was a hunch, as all the best things do, and my hunch was that plumbing and B&Q go together, and they should be synonymous. Wherever I think about plumbing, I think about B&Q. But we'd always talked about plumbing being really small category, so it didn't make any sense to me.

The data the firm had to hand was interesting in that it revealed a “brilliant” share of topic traffic:

Customers think the same as I do, and those customers were coming to us and they were all reaching products. We've not got any blockers in the middle to prevent them from getting to products, and they had a really high NPS. There's loads of them. They're landing, they're getting to products, and they're really happy, great, but it's a really small category, and our conversion rate wasn't that great. So I was like, 'I don't understand what's what's happening'.

The firm tapped into tools from Contentsquare to look at journey analysis, and started to understand that, contrary to most of its product categories, customers weren't landing on the PDP when it came to plumbing, but rather the general landing page. Worse, they were landing on the landing page, and then they were looping constantly between pages, before ending up back on the landing page an exiting. This was a problem, says Sarker:

We had a really high exit rate. That's frustration behavior. When we then layered on top customer interviews - going back to, ‘What triggered you to start that plumbing journey?’ - and everything started to make sense. Customers told us, as should really have been obvious, that nobody wakes up and goes, ‘I'd really love to do some plumbing this morning’. We tend to wake up and go, ‘Oh God, something's broken’. There might be a leak, the house might be wet, you might be wet. You might have to turn the water off, and it's stress, and none of us are plumbers.

What you don't do is what you do in every other industry where you go, ‘I want some shoes. I will search for shoes. I know what shoes I like’. You go, ‘Oh, I've got a thing. What is this thing?’. So [customers] couldn’t land on the PDP. They didn't know the name[of the thing they were looking for]. What they did was they went ‘B&Q plumbing’, landed on the landing page, and they were using the landing page to look at the pictures, to understand what it was they needed to buy. That looping behavior wasn't frustration. The need that they had in that moment, was, ‘I need to know what this is’.  The next thing they told us is next day delivery doesn't cut it when your water's turned off, so they had no intention of buying online.

The fix for this was what might have seemed counter-intuitive for most other categories, recalls Sarker:

We moved the categories down on the page, and at the top we put educational content, like ‘These are the plumbing fittings, this is what they do, this is what they look like’. And then we up-weighted click-and-collect, so we could say to the customer, ‘Now you know what you want to buy, so I've reduced that looping. But also I'm going to make it quicker for you in store, so within 15 minutes, I'm going to have that plumbing fitting ready at the front’. So, rather than arriving in a B&Q store and you've got to go and find it on the shelf, it's right there for you.

With those changes, we increased our organic impressions by 47% We decreased our exit rate by seven percent, decreased bounce rate by 38%, and we increased cart additions by one percent. A one percent addition doesn't sound like a lot, but if you're doing the maths, you're increasing traffic by that much in the upper funnel, that’s a significant change to our plumbing category.

AI, of course

As for the future, B&Q is “very much talking AI, as is everybody today”, says Sarker. B&Q has recently taken part in a beta trial of the new agentic AI Sense Analyst offering which Sarker says has been a positive experience:

What's been really interesting is watching the team and how they've grown during the time that we've had it. What I've seen is that the team started with really, really short things, and what we've really learned is the power of the prompt. So we've seen it doing amazing things, like, 'How many times have you ever lost a mapping?'. I'm always forgetting the name of my mappings. It can find them for you, which is brilliant. It can also do really basic things, like I wanted to know the other day how many people were interacting with the imagery on my PDP (Product Detail Page), so I just asked it, and it told me.

But that is the tip of the iceberg, because some of the work that we've been doing recently has been around, how can we change what Monday morning looks like? If any of your teams are like ours, they'll come in on a Monday morning, they're doing their Monday morning report, and they'll probably spend most of their morning crafting what happened last week. We're creating a prompt now which we're giving a really good sense of who it is that's asking, "What are the mappings to use? Which are the pages that you care about for your role specifically?'. And then, we're saying, 'Tell me all these different things I want to know'.

What it's producing is a dashboard, but it's huge. It's hours and hours worth of work. It's more work than the team would have done if they were building it manually. What it does is it tells them every single thing that's happening in their category, and it also gives them ideas for what they should do to make that better this week. So, Monday morning reporting has gone to hours of work to you're going to push a button, and two minutes later you've got all your answers. So you now spend your morning on, 'What am I going to do?’.

And there’s more AI to come inevitably, with Sarker taking a customer-first approach to her thinking on the subject:

We've recently launched a new AI chatbot. The approach that we've taken with it is to start with a customer problem. Similar to that plumbing journey, the challenge that customers have around ‘I don't know what the thing is that I'm trying to buy’, and how that manifests itself on our website is that traditional search doesn't work because it relies on keywords, and if you don't know the name of the thing, you don't have the keyword. So Hello B&Q is a conversational bot. It will have a conversation with you about what it is that you were looking to buy, and it will definitely help you to understand what the name of the thing is.

But more than that, it will also help you with your DIY project. So you can just say to it, ‘I'm thinking of tiling my bathroom. I have no idea where to start’, and it will ask you some really good questions, help you choose the tiles. But it will also talk to you about [how] you’re going to need adhesive, you might need grout, you might need this that or the other, and it will build up that project with you, with the step-by-steps of how to do it. So it's kind of bridging that gap where online, versus in-stores, you can't talk to a colleague. Well, with this colleague, you can talk to it, and it will be in your house when you are actually laying the tiles.