Speaker 1: Hello, folks. This is the 10-minute webinar live at 3 p.m. on the 16th of December. If you're watching this recorded, where were you in December? This is the five secrets of data-driven decision making. I hope you enjoy it. Here is Kevin. Kevin has just been shown this spreadsheet and asked what he thinks. Kevin scratches his head and said, I'm an accountant. What do you want me to do with this? Oh, sorry, Kevin. We're going to give you some more data. It happens to be the months of the year and totals at the bottom. Uh-huh. That's fabulous, says Kevin. Still, what would you like me to do with this? Oh, Kevin, it's a profit and loss account. Yes. What would you like me to do with this? But, Kevin, there's lots of data here. Can you not tell us stuff? Yes, I can. I can tell you the stuff you already know, how profitable the business is. Then comes Sarah, his manager, and says, Kevin, we're thinking of acquiring this business. Strategically, this is important to us. If we were to merge this business with ours, could you find a way in which you think that this business that we could acquire could easily slip into us and we could make it profitable without changing too much of its operations? Kevin immediately springs to life. He says, of course I can. The salary costs would reduce greatly, and I've noticed in December their sales are lower than we would have, and I would predict them to be higher. If we were to pay this value for the business, I think we could regain it within 12 months, and it could become a good profit center in the future. The point of this is that it took Sarah, the manager, to be able to ask strategic questions from Kevin before Kevin could find value in the data. When we read the media at the minute, we think big data is just simply going to resolve and solve everything for us. We look at the data, and out pops answers. Without questions, there are no answers. The question must precede the analysis of the data, and the questions typically come from the strategy planning process. This is the Ionology Digital Strategy Quadrant. There are four key quadrants. We're not going to go into the detail, but where a business resides, each of these quads has a different growth profile. The growth profile for advocacy is the slowest growth. The next highest growth is attention, but it will cost us cash. As we're paying the cash, we get attention. As we stop, it goes down again. Most businesses would like to be the authority, and they say, but we are the authority. But when we look at the data, are we? We're going to take a look at that now. The prime player is the biggest player in the marketplace, who typically just gets disproportionate amounts of attention because they are the biggest in the marketplace. Let me go back to the quadrant. The organisation says that we're growing through advocacy. It's small. We want to use digital to leverage getting into new markets, and to do that, we like that idea of being an authority. Well, if we want to go from advocacy to authority, we need innovation. To get earned media, earned media meaning people talking about us, we have to do something that will earn that media. We need to narrow our value proposition, actually say less, and be expert in a particular field in this new marketplace that we're going to go into. That's a common field, a common play within transformation. But when we take a look at what that's going to look like in data terms, we see that through advocacy, most people are coming to our website, searching for us by brand. They're looking for the About Us page and the Contact Us page, not necessarily getting too involved in our content. Our social media account is quite low. It tends to equal in and around that the amount of likes that we have equals that of the number of customers we have. Authority brands, whoever, have got a completely different profile and characteristic in their analytics. They receive lots of attention through social media. Why? Because people are sharing what they've published. They're getting lots of direct and organic referral links. Why? Because people are repeating what they're saying. They're the innovator. And that earned media is a really strong indicator that we're moving into an authority position. So a play between one strategic position and another, the data really starts to come to life and help us plot our journey and help us see if we're getting there. Of course, attention is going to have a different data profile. The data profile around that is at the cost per click, the cost per acquisition. The advertising costs are going to become evident. The new versus recurring returning visitors becomes very interesting. And prime players tend to just get huge volumes or numbers in terms of market share. The point of all of this is when we start to ask strategic questions, then we can use the data to much greater effect. The most important questions are created during strategy process. Now, I've mentioned this before, agile, lean sprints are not strategy design processes. The strategy where the questions are asked comes before them. Agile and lean and sprint and other methodologies are used to answer the questions. This is Jamie. He owns a company called Wild Goose Studios in the beautiful town of Kinsale in the beautiful county of Cork in the Republic of Ireland. Jamie's company since 1970 have made these bronze sculptures that sell very well. They've got a very strong brand. And if you're ever in an Irish tourism shop, you'll see these sitting there. Jamie then decides that he would like to do more with the business and we do a little bit of data research. The data research says for Irish gifts, there are 4,400 people looking for it. If we take a look at Celtic crosses, there's 50,000 people looking for it. And we can see all of these opportunities coming up. Jamie says, well, what happens if I were to create products that fitted these opportunities? Before he makes the product, he mocks up the product, the 3D render, one-off versions, puts them online and tries to practically give them away. The cost of the attention, using an attention model, is higher than the transactional value of the deal. Into the US, he tries to sell these for $5. Do they sell? No. Why? What the people who are typing in Irish gifts are looking for is novelty gifts, not high-quality gifts regardless of the price. So what Jamie has been able to do is to, without changing his entire business structure, without creating new molds for new products, without having to turn his entire business into something that was never going to make profit, he was able to take actions from the data that came from the outcomes. And data that influences actions is much more important than data that simply informs. So if we're going to create a campaign or if we're going to create data and start to use it in our business, it's data that allows us to say, we're not going to do that, we're going to do this. Which brings me on to number four. And number four is the juice isn't always worth the squeeze. So what do we mean by that? Well, this is what we mean. There is a tendency for those that are using data to do what are referred to as A-B split tests. In the absence of quality strategy, real understanding of strategy and the strategic challenges, many digital marketers in particular run split tests like this. I call them the insignificance test, and I'll show you why. Should the button be blue and say, click here, or should it say, learn more, and maybe should it be red? Let's test which of these things is going to create the greatest amount of sale. Well, actually, if we take a look at mathematics and what happens in these tests, binomial theory or Brunelli trials, we will actually see, and I'm not asking you to understand it by the way, but we will actually see that unless there is a huge significance in the difference of outcomes that these two tests will make, the tests themselves are insignificant. Let me show you what I mean. This is a website called Test Significance, and you can see that if I want a 3% conversion rate and I then want to try button A and it gets a 3% conversion, and button B if it moves to 3.4% and I want to be 90% certain, I actually need to get 20,000 people through my website to simply test that. Very few people have that level of volume of traffic. If I reduce it to 80%, I still find that I need 12,000 people assuming I can get a 0.4% shift from 3 to 3.4. In my experience, changing button colours, it doesn't even get as high as that. We could spend weeks of traffic to test something that yields very little and creates no significance within our site. This is fine for Amazons and Googles and many other businesses that have large traffic volumes and are really getting down into the minutia of what makes a difference, but for 90% of businesses, in particular those in transformation, there are so many more significant questions to be asked that that's where we should be doing our test. This is the reason why if you get small variances in tests, the traffic volumes required to actually test them to anything above 50-50, which was the guess at the point of where you started, requires massive traffic volumes, something most of us don't have. What we're really looking for is something that gives great, large variances. How would we do that? We would test transformation value propositions, so all of a sudden we were doing X, now we're going to be doing Y. We are the inventor of the most advanced whatever service product in the world. Test that like Jamie tested his, and you will see if you try two or three of these, you get a 1%, a 5%, and 10% conversion difference. You don't need large traffic volumes for those, and they massively increase the perpetuation of transformation within an organisation. Here is number five, the last one, graphs that never go down ruin your reputation. I have never seen a graph for social media likes or following go down. 14% of marketers are at the boardroom table in US businesses. This is typically because they measure things that they like, that they want, but have no significance within the business whatsoever. The data that matters in the business is that that aligns with the business objectives. If marketers are unable to understand the business objectives, we have a leadership problem. Inside Google Analytics, we can set up our business objectives, and we can measure them. This is an e-commerce one, it does not have to be e-commerce, it works equally well for business to business, but there's a thing called attribution modelling. We can check along the way when someone went to social media, through email marketing, through organic search, and came back, and then eventually downloaded a brochure, created a phone call, filled out a feedback form. We can test how the components that made that customer journey actually worked, and this data is actionable. This is why we need to move to actionable data. Graphs that never go down ruin your reputation. Data that enhances your reputation is that that leads to business objectives. You can sit at a table with finance directors and chief executives and talk about sales and marketing, customer conversion, customer acquisition. This is where the game is at. Not only just graphs that tell what made them acquire, data points that allow us to tweak it and change that outcome, enhance that outcome. Folks, I thank you for watching, and I wish you very well. Thank you for joining me on this presentation, and I'll see you on the next webinar. Thanks again.