Blog

The hunt for a bra that doesn’t pinch, slip or create unflattering bulges can be long and frustrating. That makes choice really important to shoppers seeking comfort, support and hopefully, style.

Online lingerie and swimwear company Figleaves prides itself on offering plenty of choice – but that also leads to it having to manage a mountain of data about its stock.

A bra from its own-brand product line-up, for example, can come in 36 different sizes and five different colorways. So that’s around 200 individual stock keeping units (SKUs) for that one product alone. Factor in all the products it sells from some 100 other brands, including Freya, Fantasie and Chantelle, and you’re looking at some very complex data analysis problems.

For this reason, it’s perhaps unsurprising there are questions that Figleaves wants to ask about its business that were, until recently, quite hard to answer. For example, Angus Jenkins, the company’s head of e-commerce, has been keen to explore how online advertising spend translates into profitability across this product portfolio. As he explains:

I can look at our advertising through the lens of the demand that it’s driven, in terms of gross revenue. And marketing spend seen in terms of gross revenue numbers is certainly something that tells us whether we’re doing a good job or not. But what we’ve not been looking at is the margins we’re actually driving. So what’s the gross profit or contribution against a particular marketing spend?

Getting answers to more in-depth questions, he says, often means bringing in external data, from sources such as Google (from its AdWords and Analytics services, for example) and Facebook, and integrating it with the company’s already complex data mix. Going further still, to integrate details of competitors’ pricing for the same and similar products, meanwhile, will give Figleaves a way to discount more dynamically while still keeping a careful eye on margin protection.

Marketing optimization

This kind of marketing optimization is a big challenge, but it’s important in terms of strategy, says Jenkins. Figleaves has brought together marketing and merchandising employees as a core e-commerce trading team, and is getting them to work much more closely together. A big focus for this team is actively exploring how strategies to both attract customers to the site (marketing) and entice them into making purchases (merchandising) might best be combined to result in the most margin-rich sales. Says Jenkins:

We see a lot of opportunities to be had from greater insight, from just making sure we’ve got the right amount of stock in the right colors and sizes to deliver a profitable business, to adding customers segmentation data to better understand how different customer types interact brands and categories, depending perhaps on their age or whether they’re in the UK or US. And as we increasingly explore these areas, I expect we’ll come up with new areas to explore.

With this in mind, Figleaves has recently implemented an analytics platform from Avora to help it fulfill the business intelligence needs of this trading team. Avora is a London, UK-based start-up that uses machine learning to create ‘smart alerts’ to changes in data and root-cause analysis to suggest to users the underlying reasons for these changes. The idea here is that this enables companies to sidestep some of the legwork typically associated with analytics.

Earlier this year, it announced $6.5m (£5.1m) in Series A funding, from venture capital firms AlbionVC and Crane Venture Partners, as well as angel investors including former Salesforce EMEA chairman Dr Steve Garnett and former Sage CEO Stephen Kelly. As well as Figleaves, other Avora customers include telco O2, online retailer Ocado and news organization CNBC.

Avora’s integration strategy was a big selling point for Figleaves, according to Figleaves head of technology Graham Smith. The company offers some 350 pre-built connectors for different data sources, including internal systems that Figleaves uses, such as Salesforce, as well as many of the external sources associated with its online advertising campaigns. Says Smith:

The links we would need to our various data sources came up time and time again as a significant development effort and an ongoing overhead for this project. That integration work – well, it just didn’t seem like there was any value there in us taking the work on ourselves, because it’s just about getting data from one place to another. So the general idea of a pre-built connector, managed by the provider, had a great deal of appeal – as did the proposition of getting off our reliance on on-premise SQL Server for our analytics needs, and handing the management to experts, too.

Up and running

Speaking to diginomica at the end of August, Jenkins and Smith confirmed that the Avora platform is now up and running and that Figleaves is now working through a long list of use cases that it wants to deploy on the platform. The biggest challenges, according to Smith, have not been associated with the technology deployment itself, but at better understanding Figleaves’ own data – and in particular, the different definitions of that data among different departments.

But as these issues get sorted, Jenkins is optimistic that what Figleaves will end up with is an analytics environment that encourages members of the e-commerce trading team to experiment with and explore data, rather than depending solely on reports and dashboards:

We’ve developed the trading team to be self-organising and proactive in terms of going looking for insights, and identifying the correlations needed to make good decisions. So what we needed was a system that would support that, which would help us get to insight quickly by bringing together lots of disparate bits of data and sticking it somewhere where it could be interrogated and explored.�

This idea of a ‘journey of data exploration’ was pretty critical to us, because we’ve got really smart, skilled people who need more than the quite static reports and dashboards that other suppliers provide and which are hard to take further without quite a lot of investment and upskilling.