Garry Viner: Hello, this is Gary from Rocket Agency. I’m sitting here with David Lawrence, Managing Director.
David Lawrence: Hey Gaz.
Garry Viner: Hello. You have written an article this week, potentially controversial, entitled Just Because it’s Data, Doesn’t Mean it’s Right. Why are you, why are you thinking this way?
David Lawrence: I’d love to think it’s controversial. Sadly, I’m going to burst that bubble straight away and say, I really don’t think it is. As an agency, we have tonnes of data that we get our fingertips on all the time, and we love data. We make some incredible decisions with it. But something that really worries me is when a client comes in and they automatically say, we tend to make all that decisions based on data.
And for so many of our clients that’s just not possible.
Garry Viner: Yeah.
David Lawrence: And so it worries me. And I wanted, I guess, to start the conversation around the fact that just because you have numbers in front of you, it’s reckless to assume those numbers are correct or will even lead you to a conclusion.
So that’s why I wrote the article, and that’s the conversation I think we should have.
Garry Viner: Okay, so those clients for whom a data first approach is not always the best way, what are the commonalities? Who should be mindful of the fact that data is not necessarily the right way for them to go?
David Lawrence: Absolutely, I think everyone should be mindful of it. I mean, probably one of the really common real world examples that I see a lot is when you have a client, often a B2B client, but any client with a long sales cycle, we see clients where they look at metrics in Google Analytics and they’re looking for those metrics to be signposts pointing them in the right direction. And so often the reality for them is that they have a very small number of high value sales per month.
Garry Viner: Right.
David Lawrence: And they’d be much better off looking into what those small number of people did rather than looking at the masses of data in Google Analytics. They might find that their actual customers that create value for them are 0.1 of a percent of all the people that experienced their website. So if you’re looking at those big numbers in the website, you’re potentially being led into decisions of people that weren’t even worth money to in the first place.
I know that I probably look at data from more of a usability point of view than you do. You get much more in the actual numbers. And I know you’re an actuary by trade, so you obviously love the numbers and the data. So how about you share something that you come across in terms of data limitations and problems?
Garry Viner: Yeah, I would say that these, again, aren’t limitations for the data, but it is a limitation of how data gets used. So I can think of a couple of examples. So first one is, for data to be meaningful, you have to have, a critical mass. You have to have a threshold of data at which it becomes statistically significant. And the unfortunate reality is that for a lot of clients, they just don’t have that level of data. Or if they did have that data, they would have to wait potentially months to see a data that was meaningful for their business. And so the very idea that you could have a low traffic website and draw conclusions of that in a month or even three months, sometimes that just doesn’t happen.
The other issue I guess, is that data is only as good as the inputs. And so often we will see misconfigured implementations. I can they have a couple of examples. Google Analytics having been misconfigured in such a way that attribution at the channel level is not correct. And so you might be tempted as a business owner to take money out of AdWords for example, or SEO because you don’t think that’s driving conversions. And in fact it’s a simple configuration issue.
And another example recently where a revenue tracking was not correct, and this is interesting because it’s not just humans that are prone to making bad decisions. Artificial intelligence is also very much a function of the data inputs. And in this case with revenue tracking incorrect, how we saw some campaigns that were optimised for revenue, and of course they failed because it just wasn’t set up the right way.
David Lawrence: Interesting.
Garry Viner: Yeah, so it was actually a very simple thing to kind of fix up the campaigns that had been struggling for months and months and now they were running very well. But had that approach been taken right from the start, nothing would have gone wrong in the first place.
Dave, what tips do you have? I should also say, by the way, that there’s plenty of good data that we take out of Google Analytics.
David Lawrence: Plenty of good data.
Garry Viner: Plenty good data. It happens all the time. It’s most of what we do. Looking at ads that are getting best results, looking at keywords that are getting best results, optimising towards them. It’s a no-brainer.
David Lawrence: Yeah.
Garry Viner: These are the minority of cases that we’re talking about here.
David Lawrence: Yeah.
Garry Viner: You just have to be aware of them.
David Lawrence: We have to be aware of them, that’s right, and I think something that we find as well is just because a company is large and has sophisticated resources in-house, doesn’t mean their data’s going to be right. I often find that the larger company gets, the more spread out their product and their data sets become as well and you can always have more errors.
Garry Viner: Yep.
David Lawrence: But definitely on that point of being sort of positive about data, absolutely. Tonnes of great decisions. And we had a really good one the other day where we had a client came in, we’re doing keyword research. They saw things in bulk, and our research indicated that there was a lot of traffic around for their particular product and the word wholesale. The client weren’t really keen on that because technically they’re not a wholesaler and we understand that. But we said, “Look, bear with us. Let’s go with what the data is suggesting and run it.”
And we did. And it’s been actually one of their best performing campaigns. And the reason is that even though they’re not a wholesaler, people don’t necessarily understand that. They interchangeably use the word wholesale and bulk. So if it wasn’t for looking at the data, if we’d been more relying on the reality, we probably wouldn’t have come up with that particular keyword. That campaign would have suffered as a result.
So, I guess in the spirit of helping people get to a better place with data, if you had a particular tip that you would suggest people do with their data to make better decisions?
Garry Viner: Yeah, I think just making sure that your implementation is kind of double checked right from the start. Having reasonable expectations in terms of time frames, being prepared to think behaviorally as well as kind of econometrically, so to coin the phrase.
David Lawrence: Okay, actuarial term?
Garry Viner: Actuarial term.
David Lawrence: Nice one.
Garry Viner: But obviously, where the data is kind of meaningful, then you want to follow that.
David Lawrence: Yeah, I like it. I’ve got probably the one for me is around the business side of things. One of my pet hates with data is when people go into Google Analytics and just desperately look for a story. So I would always suggest that people have a question, have a business question, they want answered and they go into Analytics or whatever their package is to find the answer to that.
Otherwise, what I think happens a lot is people go into it, they look for a story that maybe doesn’t exist, and they focus on vanity metrics, which may or may be important for them. So bounce rates, time on particular pages, path through the site.
Garry Viner: Yep.
David Lawrence: Stay right away from that. Go into Analytics with a clear business question and then make decisions based on that.
Garry Viner: Brilliant. Okay, and any final message for yours?
David Lawrence: Look, I think if you have problems with data, we are always really happy to have a talk about it. So there’s lots of more resources on our website, but got any big questions, big problems, get in touch with us. I’d love to have a chat.
Garry Viner: Thanks so much, Dave.
David Lawrence: Take care.