Lucia says: “If you want to build a successful campaign, you often think that, from your SEO point of view, it's gone really, really well – but, actually, you might receive an underwhelming response from your client. In that case, that's when this has happened, and there's been a break in the failure cycle.

So, as an SEO, you're measuring all your things – you're measuring hits on the website, you're measuring the organic traffic, you're measuring engagement, and eventually the sales. But, from a client's point of view, as well as measuring acquisition (which is very important), they're also looking at what's happened on their products, which products are sold, which ones have been returned, stock optimisation, order values, etc. They're looking at clients that repeat buy again, and they’re also understanding their margins and their profits.

If, for example, there's a Black Friday campaign that's been run and, from an SEO point of view, it's been very successful. But by the time you get to January, the client's seen there's been a lot of refunds happening and they haven't actually achieved what they wanted to. They're going to end up with lower profit than they expected and they're going to end up with their cashflow not being as good as they wanted it to be.

When you're not looking at things the same way as the client, this occurs. The way to really overcome it is to work very closely with the client. Look at their specific goals, try and understand them, and make sure that you’re working on those – whether they want to get rid of some stock that's not performing very well, or whether it's the fact that the margins they need to achieve aren't being done.

When you do that, you can know it and be more successful and more effective when you run your campaigns.”

How often do you need to align your KPIs with your clients? Is it just when you're onboarding your client and getting to understand their business, or do you need to do that on an ongoing basis?

“I think it’s on an ongoing basis. Your client will have primary goals that they will tell you for the campaign, for what they want to achieve, but they'll also have secondary goals in their business. These secondary goals will be other things they want to achieve. Over time, that will change as their business develops, when they've got specific things in mind, and as the environment moves.

The more and more you can work with them, constantly refine it, look at what they're doing, and try to help them achieve it – the better you can go.

Their KPIs will also change over time as well, depending on their particular business needs. If they introduce a new item of stock or a new product line that suddenly starts performing well, they might want more focus on that and less on other things. Or it might be that they're running a sale and their margins start dropping.

There's a constantly moving picture and we can't assume that it's going to stand still. The more we can understand what they're looking at, the more we can help them.”

What is the ‘built-in failure cycle’ and how do you go about breaking that?

“To me, the built-in failure cycle is the fact that you've got a campaign going on and, when you run a campaign as a marketing or SEO agency, you look at all the marketing data. Then, the retailers look at what's happening on their e-commerce platform in terms of their product sales and customer retention. They're looking at margins that are being achieved. Then the accountant is also producing some profit and loss reports.

3 sets of reports are happening, and they're also happening at different time periods. The campaign stats for the time period of the campaign, the ongoing activity through the sales on the e-commerce platform, and then this different measure that the accountant's working on. You're trying to look at 3 different types of data in 3 different reporting periods. That's where the failure sets in because you're looking at different things.

If you're only looking at the acquisition part, you're not seeing the rest of the picture and understanding the whole view of the client.”

Is it possible to combine those data sources, or those types of reporting, so that everyone's looking at the same thing?

“It's possible to do it. It's not an easy thing to do but, if you've got the right skills in place, you can then get the API that takes the information and brings it all together. The way my company is working is we're building a data warehouse. What we do is we then take in the information using the client APIs off the e-commerce platforms and, within that, we can pick up some of the marketing data.

If we've got the most important marketing data there, we can then pick up all the activity on the platform and then look at the profit margins for individual products, or the daily profit on the sales.

Then we feed that into a data dashboard. Within this data dashboard view, we can then align things and set filters there. The larger companies that work in the retail environment have had access to this tech for years. But, for the smaller SMEs, they've had barriers there (time, budget, etc.) when trying to develop the project.

I think solutions are out there, but a lot of the solutions on the market tend to look at 1 problem in the market – whether it's attribution or stock. Very few of the solutions out there have these extra add-ons to look at the whole picture. I think that's the gap in the market.”

Do you attempt to have meetings with accountants within an organisation to understand how they report upon things – and serve what they're looking for as well?

“Within the retail space, using the kind of data reports I've talked about, you can understand how much profit you're making on each day of sales. Then, from the accounting point of view, you know what your fixed costs are.

If the accountant can tell you what the fixed costs for the month are, then you know you need to cost that amount of money, and you can then start getting those ratios in there. You don't need to take away what the accountant does, but you can provide them with the information they need.

They're also very interested in how the stock has moved over time. Especially for small retailers, for them to count their stock every day of every month, it's just painful to do that. We can also help just by providing some of that information and joining it more closely together so the accountant can see what's happening.

Accountants often see the big picture, but they can't go into the details. They don't know which products were the most profitable, they can't see the concentrations. Was it 10% of the customers that brought in half the value for sales? Was it the fact that 1 particular product line is always the best seller, but then it's got a low margin? If they could see that information, there's definitely potential to work far more closely together.”

Are there any other departments within a business that you'd like to try and bring in and understand?

“I really like to try and work with everyone. If you can get a good board pack – it's all about having the board pack. It’s about getting the whole picture, making that same set of information available to everybody, and really trying to get that information as quickly as it can be made.

I don't want the business to spend their time trying to add up numbers and constantly do something manually because, when you do things manually, it's easy to make a mistake. It's easy to make errors in it.

Whereas, if you can get to 1 dashboard that everybody understands, you can then give different people in the business different access. Your marketing people might get access to the marketing campaigns part of it. You might want the financial director to see the financial information. It really is about getting that information right at the right levels.

If you've got 1 particular team involved in 1 particular product – for example, if you've got a retention-focused team, then obviously you want to give them access to everything that's around retention, so they can then see how these customers go on to buy in the future months.”

How do you incorporate more of this into the way that SEO is done, or even into your SEO training?

“It's quite a new concept for many SEOs because, often, they've never had access to this information. Although they're very good at what they're doing, they're not seeing the bigger picture. It really starts with some kind of training sessions, bringing the awareness that all this data is out there and it can be used by you.

Then it's about getting their data into the system and starting to customise it so it makes sense to them, and so they can see the world in the way they see it – that also makes sense to the retailer.

It's really about trying to bring those 2 worlds together and, if you can do some training where you can say, ‘Here are some examples of case studies showing the way people have worked and how they've changed their working practices.’

The excitement is that, once the data goes into the system, you suddenly think, ‘Yep, here's the big picture.’ Then I'm interested in breaking it down and looking at this query, this way of doing it, or this particular filter. When you can then build that, you can then bring the world to light. It suddenly makes sense to people. They get this light bulb moment: ‘Aha! I can see this, and I can see how it's impacting that.”

Is there any particular software that you recommend for data warehousing?

“There's a lot of different software out there, but I think the skillset isn't really about the software – it's about knowing where you want to get to. In your head, if you have an idea of, ‘The data sits here. This is the point I want to get to. How can I bring that information together?’ Power BI is one that's commonly used. There's everything in Looker Studio through Google. We like to use one, which is another party called GoodData, but the e-commerce data is hard to use. It's not friendly, easy-to-use data. It's very complicated.

For every product that's set up, there are product children beneath it. If you want to look at customer lifetime values, you literally need to look at the first time a customer purchases. Then, every time an order is made, you need to be able to link it to that customer. You also want to check for errors, so that there are no problems with the data coming in and it all makes sense. Then you also want to be able to add it and combine it.

There's definitely a tech element to that, and that's a real barrier for SEOs to actually use the data. It's not simple data. It's complicated to use and it also gets large very quickly. If you're going to start with every single order ever made, from the time the company was set up, you've got to be able to work with a big, big data set. I certainly have a data developer who's been really helpful in supporting me for that part of it.”

Are you a fan of using AI to assist you with crunching data?

“I think it's got its place in that. What I'm most interested in is really getting these basic fundamentals out. Once the fundamentals are out, then companies can work out how to use it, and then work out how they use AI. I don't think AI on its own is the solution.

If all the data is there and available to you, you can then say to AI, ‘Tell me this part.’ If you want to look at a new product that's performing well, you can then get AI to focus on that part of it. But I don't think it's got the complexity to build the thing in the first place.

It's really about having that vision of what you want to do, how you're going to achieve it, and then putting AI in to answer specific problems. Those specific problems might be relevant to the business but, if you're focusing on one part of the business, then you might miss the bigger picture.

If you can access everything straight away, then AI can help you delve into those really juicy parts where the opportunities are for your particular business at that moment. Understand your client’s goals, to begin with, work out the systems that you need to achieve that, and then perhaps look to automate that in the future – where you know, specifically, that there’s a business need to do that, and a benefit that's going to be there.”

What sort of campaigns are you finding that delight clients nowadays?

“One example was working with a customer who was introducing a new helmet range, and they bought the same helmets in 3 different colours. They put in the same marketing budget, and they cost the same. We then looked at the sales, and how they're developing over time. What was really interesting was that, of these 3 helmets, one of them performed amazingly – it got really good sales and had low refund rates. Another one was doing okay, and the third colour just wasn't as popular.

By being able to distinguish that the third colour wasn't as popular (the sales were lower and then the refund rates were higher on it), we then looked at the stock in the warehouse. We could see that, at that point, the stock was starting to build up. We suddenly realised, forecasting forward 18 months, that this one particular helmet was going to become a real problem. The money we'd wanted to get back in wouldn't be there because it was all sitting in this helmet that hadn't sold so well.

What we actually managed to do was change the marketing budget, put out a lot more blogs on this particular colour, reduce the marketing budget on the more popular colour, and then also put some sales and some pricing incentives in. Within that campaign for this particular retailer, we ended with a campaign that cost more money and brought in less sales, but we weren't then left with all the stock in this unpopular colour.

The outcome of the campaign was that we actually achieved the retailer’s objectives, but not necessarily in a typical marketing way. I think that's a real interaction. When you manage to interact all these things together, you can actually carry out a campaign which makes the retailer happy, but doesn't necessarily have only the marketing goal as the main focus.”

These aren’t necessarily conventional SEO metrics that you’re talking about. Are these metrics the future of SEO? What are the key SEO metrics that you like to recommend nowadays?

“I haven't come from an SEO background, I'm not an SEO person, and I really only moved into e-commerce after the lockdown. I'm coming from a data background. My background is that I spent years and years in big banks and other organisations, and I worked a lot with SMEs. I come from a different perspective.

From my point of view, I don't think the SEO metrics are enough on their own. They need to be able to go all the way in. We can take the organic traffic on the website and then we can actually look at that to work out who the customers were, what the retention rate was for those customers, what products sold, what the refund rates were, what the daily profit and sales that brought in, etc. You can then compare your organic traffic against the paid traffic and look at the whole picture of the sales.

For me, that's the market – saying, ‘The SEO impact was x compared to the paid traffic impact. I like to look at everything altogether. As SEOs, we can track that through, but we can't get to everything. It's about actually showing the impact of the overall business, then a company can make a formal decision. Does it make sense for them to put more money into SEO? Does it make sense for them to put money into paid advertising?”

What's something that's seductive in terms of time, but ultimately counterproductive, that SEO shouldn't be doing in 2024?

“I always look at attribution. In my data analyst head, I know that there's ambiguity there. We know that attribution is very complex. We know that people go from one site to another. They look at different channels. We can never get to the full picture. I think we have to embrace this and be clear with our clients.

We always know there's some level of ambiguity. With the best will in the world, we'll never get to a complete picture. Sometimes you need to be proportionate. If your client hasn't got a very complex structure, and you can work out what the gaps are and what you're not seeing, just work with those assumptions and gaps.

You know it's proportionate, so you can proportion it out into measuring your SEO and the ambiguity there – but you don't want to overspend on it because you might be better off putting your time and budget into doing more marketing or more campaigns, rather than really trying to track down and solve every gap.

Whatever gap we target, we know it's Whac-A-Mole. Something new comes along and there's something else we can't measure or can't track. Don't go down a rabbit hole. Just be aware that attribution is always going to be a complex problem and things will happen. One day sales will be twice what they are on another day, and we don’t really know why that is.”

I don't hear too much about attribution models nowadays. Why is that?

“I think the software has come in. Some people are really, really into using the attribution software and they're going to use it. For other people, the market has moved on now to AI. AI is hitting the attribution market, I think.

There's always something new and exciting in an industry. At the moment, everyone's talking about AI, which is why the focus is turning away from attribution.”

Lucia Dello is the Founder of Dello Insights, and you can find her over at DelloInsights.com.