[MUSIC PLAYING][Business Statistics]I'm Stuart Duncan.[Stuart Duncan, Director of Data Science, Simply Business]I head the Data Sciences team here at Simply Business.I've been working with the company for abouta year and a half now.We're a company that does online insurance for small businesses.

[High Street Tracker]The High Street Tracker is a piece of analysiswe launch and release every year.What it is is information about movements on the High Street.We collect-- as people come on our siteand fill in the request for quote processand go through the process of getting an insurance quote,we collect a lot of information about what they

do, the industry they work in.So we aggregate that up.It's a useful set of information to help understand movementsthat are happening in the small business sector in the UK.And we use it to improve our servicesthat we offer to our customers and to help the waysthat we communicate with them and understandingof our market.[Using Big Data]

We use big data technologies to help us manage and trackour data that we get.The world's changing, with more data sourcesavailable than there have been in the past.So you need different technologiesand approaches to deal with that, to make it into somethinguseful.I'd say big data is a little bit dependent on your industry.Certainly in the online space where

there's a lot of traffic to a website,it's very important to be able to collect and scalethe collection of that information,and to be able to also crunch it and take all the raw dataand turn it into something useful, a nugget of informationthat you can use.[How it Works]So the Tracker itself is an output of informationthat we store in a database that we keep.

It's actually in practice a series of databasesthat we have that stores information around the quoteprocess and how people are answering questionsas they go through the quote.So essentially when you're on our siteand you kind of come over here and youwant to start a quote process, youclick the Compare Quotes button.And this is effectively how you go

about requesting a series of quotes for insurancefor your business.And you would sort of fill in a numberof different effectively data points over here.And these data points would allowus to generate a series of quotesthat we could then provide back to you that you could thenpurchase.So these data points will go into an engine.That engine will figure out a quote price

against a number of insurers.That will return a set of results.That results will be like-- we call it a panel effectively,a panel of results.And what that will then do is allowyou to go in and buy the insurance itself.So that information that's going through this processthat our customers are providing iswhat goes into our database behind the scenes.So apart from allowing us to give you a quote,

it also allows us to aggregate and understandthe information our customers are providing.So in a sense the Tracker itself is really a piece of analysisthat we do on top of that data, where we'retrying to answer questions.For example this question, how many yearshave you been operating your business for,as that gets provided to us, we canstart to see over time whether we're

getting more companies in the economy that are olderor younger or starting up or, you know, kind of justbeginning.[How Data is Collected]The data we collect comes from actually a numberof different places.The primary place we get it is from the customer's completingthe quote forms themselves on our website.

But we do get other bits of data from what they do on our siteas they go through that process.Those questions are determined by the insurer themselvesand how they want to understand the customer's risk.As they go through the site, they answer those questions.There are other ways.So they could also call us.We have a call center too.

And in that process, an agent wouldwalk through a similar form with themand collect that information.We're able to see how they-- what kind of pages they view.We have a big knowledge community,which gives insight and kind of ideas around howto run your business.So from that, we're also able to see what pieces of contentthey're interested in.So we collect that together into a central database.

In terms of the High Street Trackerinformation that we have, we have close to 300,000active customers.But we have far more information through the quote process,so it's in the region of millionsof sets of different bits of data that we collect.[Analysis]

So the database behind the scenesis really where we store information.And we have a number of tools that allowus to access and review that.So we would go through and organize our datainto different dimensions and categories thatallow us to interact with it very easily.

So some of the core measures that wewould have would be visitors that we receive,things like the number of policiesthat we've sold, prospects who'vebeen requesting information.And then we keep information against their trade,which would include things like the shop,if they're a shop, the type of shop that they are and so on.

We also look at understanding where people come to us, howthey arrive on our site.So that's what we think of as our marketing channel.Are they coming to us from a paid search ad on Google,or they're coming to us through an inbound call.We use that to understand how better to communicate with themand how better to attract other customers like them.What we'll do is we will analyze and look

at trends or interesting observations we can take,interesting bits of insight that we can get.And then depending on what that's saying,we might decide to include it or exclude it.Generally we're looking for piecesthat are unusual that you couldn't get from other datasources out there, like government sources.Because we have a unique data set,we want to find things that are important and probably unknown

elsewhere.There's a lot of work that actuallygoes into taking the data once we've collected itand then making it relevant and usefulas insight to the different stakeholderswithin our business.I have a team who spent some timeto set up processes that automatically enrichand transform that data into something useful that

can be used by our internal customers,our internal business colleagues.So we provide that information back to themthrough a number of different tools.And they're able to go in and answer business questionsby slicing and dicing that data, looking at itthrough different visualization mechanisms.And they're trying to look for thingslike trends that might be useful in their different areas.So for example, the marketing division

might be interested to know if there'smovements happening in different regions within the UK.They want to put more attention or more focusin a particular area, so are we seeing more customerscoming to our site from maybe the north than from the south,or whatever the case may be.[Business Analytics]Business analytics in general, so the process

of collecting data and getting insight from it,is something that's used in all industries,very heavily in the online space.So where people are selling products,it's a very competitive area generallyso companies will collect information there.And they'll use to help make better decisions about howto run their business.So it's definitely not specific to insurance.I think what makes insurance interesting is that we collect

different types of data to maybe an Amazon or an eBay or a bankor any other kind of company like that.We work very hard to be what's known as a data-driven company.So we try to use information that we collectto make better decisions day by day about how we organizeand run our business.

So it's very important that every unit or every areawithin our company is actually aware of what they can--what insight they can get from the informationthat we have available.So we evangelize.My team go out and evangelize a lot.We work very closely with our own colleaguesand help them to understand what they're seeing.And we try and give them as much access

to raw or very low-level information as possible,so that they can work with it and spot the trendsand spot the opportunities.[Data Access and Application]We have various ways that our colleagues can access and usethis data.Many of them access it through Excel as a front end,as a tool to go in and do further analysis.

The database we've created is accessible inmany different types of tools.We've been using one called Tableau.It's a quite useful visualization tool.And the data itself is used throughout the businessin various operational meetings.At the moment, we are looking at running radio campaigns

across the UK to promote our services.People will enter where their businessesare, post codes, et cetera.We'll use that and aggregate thatto understand what type of customerswe have quite locally actually.We have a big focus on trying to be as local as possiblewithin the UK.So we'll take that and our marketing teamwill then look at where we're seeing

a change in a particular type of businessin general asking us for information.It might be in Birmingham or it might be Oxford, Cambridge,wherever the case may be.And then we'll use that to target our radioads at that area.So we'll understand there's a large number of plumberssearching for insurance in Oxford, so it might be a really

good idea to have a good promotion, a good kindof message to people, to plumbers in that area.And that'll help us to optimize where we try to communicateand market to our customers.Similarly for things like billboards and so forth.On the insurance side, the information that we collectis very useful for our relationshipswith our insurers, the people who are providing us

with the products that we sell.So we give them insight to help them understand whattypes of products people are lookingto buy in the insurance space.Are they mainly looking for landlord insurance?Are they mainly looking for public liability insuranceto protect their businesses against potential legal actionor something that could happen?Or is it employer's liability insurance?

So we'll help them to understand how to adjust and changetheir products to kind of reflect what the customers areasking for.[Using Data to Evolve]Generally on a Thursday like todaywe have a number of meetings in all our different teams,our acquisition team, our call center team, our managementteam.

And they will review the informationthat we have against our targets for our salesand the kind of targets we have for number of visitors.And they'll adjust their strategies for the weekbased on what they're seeing and how well we'retracking towards achieving our business objectives.The data tracker definitely informs the way weas a company evolve.It helps us understand where to focus our resources internally.

So that could be growing out our call centerif we see that we have a lot of people wanting to call us.It could be potentially starting up a new officeor something like that.We do change our internal structure at Simply Businessquite often in response to opportunitiesthat we're seeing in the marketplace.So just recently the High Street Tracker data

helped us to identify a new segment we wanted to target,so we created a team around that.We call them the Grand Designers.They're a segment of professionalsthat we hadn't expressly been targeting before.So what we do then is we create a team.We'll put some resource, such as some webdevelopers, some product people, some insurance people together.And we'll say, right.Well, how do we go and build great products

for that particular market segment?And you'll see it reflected in our track structure and the waythat our development resource getsassigned to go after particular parts of the industry.So the mechanisms and processes weuse to put together the High Street Trackeris very much a fundamental part of how we operate.

The data that we collect from that and the waythat we get that to people is front and center in everythingthat we do.And we'll keep on investing in thatand we'll keep on trying to understand customersbetter and make sure that we can give themthe best products we can.