Matt Brittin was talking to The Open University Business School's Fiona Ellis-Chadwick after a recording of The Bottom Line.
Okay. Data is driving a lot of the things that we do online. Can you tell me how important is data for the digital economy?
Matt Brittin, Google
Well one of the great benefits but challenges of the internet is the amount of additional data that is generated by people doing things online. And so what I see is when people are searching online on Google, they generate some information which we use, and the information might be I search for a digital camera in this location and then clicked on this thing at this time, that’s all we know, but that helps us improve our business. So all the results that we serve, we can see over millions of digital camera searches, which were most relevant based on what people click on, and then adjust and improve our results accordingly. And the challenge is I think out of all the data that’s generated by millions of people on billions of websites doing different things is to figure out all what’s important and how can I improve what the consumer does online as a result of it and therefore make my business more successful.
A great example of that would be retailers. If you look at a retailer, highly data driven business, a supermarket like Sainsbury’s or Tesco’s for example, lots of data generated every day by selling products in stores and are we in stock or out of stock and what’s the price point and so on and so forth. And online they're only just starting to figure out how to use the online world to generate new insights and improve their business.
So online when somebody lands on your website, they then spend some time on it, they go to view the different pages, they put something in a shopping basket and spend a certain amount of money, and if you look at those patterns of information, they can help you improve your business. You can test new price points online, very, very quickly, you can change the way the website’s laid out and so on and so forth, and in doing that improve your business. And at the moment it’s really only the internet sort of pure plays, by which I mean the eBays and Amazons of this world, who've got that online approach to data and testing right. Many traditional businesses are starting but doing many fewer tests than the pure plays.
So, for the companies that really understand the data, to what extent do you feel how we search what we search for is taking the guesswork out of advertising?
Well some of the things which we try to do is make the data that we have available available to everybody. So there’s a really interesting free tool called insights for search, if you search Google you'll find insights for search by typing that in. And what it does is it allows you to see patterns of search behaviour over time and compare different words one with another. So a nice thing to look at would be two favourite TV shows, compare Strictly versus X Factor and see the seasonality and which is rising or falling. But actually for a marketeer you can compare your brand, your products and services, your competitors’ brands over time, and you can slice and dice it by geography, by time period and so on, and very quickly that allows you to build up some insight based on people and what they're actually doing on the internet - not what they're telling you they're doing in a survey that’s several weeks old - but what they're actually doing on the internet and how that’s trending.
An example of how people can use that sort of data, imagine you're a fashion retailer and you’ve got a shop and you're not interested in selling anything online but I think that the internet can make your business better, even if you were just doing a traditional business. So if you go to insights for search and type the term maxi dress, what you'll see is it’s seasonal, but you'll also see that each season over the last few years it’s got bigger, the seasonality has lengthened and it’s become more and more popular until this year it looks like they might be fading away. I don’t know fashion but I know what the insights for search data tells me. If you're the retailer, that information could have allowed you to invest in the right stock early, to carry it for the right period of time and start selling out and not restocking at the right point in time, and each year to manage your buying to suit the customer demand better, and so that’s a good example I think of how internet data can help even traditional parts of the value chain and traditional businesses make smarter decisions.
Okay. If I was a retailer, a startup retailer, what would your top tips be for entering the online world of marketing, data analysis, what are the top tips for me?
A few things I'd say, firstly don’t worry about it being technical and dangerous, I mean actually it’s getting simpler and simpler to do things online. Over the last 18 months we've helped 200,000 businesses get online for the first time, free of charge, 20 minutes, no technical knowledge to get a website, and I think it’s just much easier than you think. The second thing I'd say is online the same rules apply as offline, in that it’s important to be clear what you stand for, why people should consider buying from you, be transparent about your prices and charges and so on and so forth.
So I see too many websites where trying to understand the cost of buying something is buried on the 19th page before the checkout, where people aren’t clear upfront what they're all about, where there aren’t photographs, you know, and photographs really help us understand what a business is all about. So just putting yourself in your customer’s shoes, go online, look around at what your competitors are doing online, look at what a good experience and a bad experience is, and apply those things.
So why am I distinctive, why would somebody want to buy with me, what does it cost, what do I look like, those are all really important things to get right. And then I think the data that you generate, you need something like a free tool like Google Analytics, which allows you to start to see well where did people come from, how long did they spend on the site, what did they do, and that helps you to start to optimise the experience for the customer.
So if, for example people come in from a Google search for maxi dresses, to your page about maxi dresses, and then a very high proportion of them leave, suggests there might be an opportunity to improve the page on maxi dresses. And I saw a great example of this, a retailer that we work with had what we call a very high bounce rate, that’s when people come and go off a page, for women’s dresses. And it turned out the reason was the default way that the dresses were ranked on the page when it first came up was by price, so actually what you got was 25 wedding dresses.
So your consumer thought she’d come to the wrong page, but the retailer had thought I'm going to rank them by price because I might sell more expensive dresses, and actually of course they want to be ranked by most popular or you want to sort of try different configurations to see what works. So that’s the kind of data that the internet can really help people with.
Fabulous, thank you very much.
Okay, thanks very much.