Computers and artificial intelligence have come along at an exponential rate over the past few decades, from being regarded as oversized adding machines to the point where they have played integral roles in some legitimately creative endeavours.
It’s a challenge many larger businesses face around digital and data analytics: How to break down organisational silos that exist in order to get teams to work more efficiently, strategically and innovatively. And it’s a challenge that was facing Coles until two very separate analytics leaders from its financial services and online businesses joined forces.
Speaking at the Adobe Symposium in Sydney this week, Coles Financial Services digital analytics and optimisation leader, Robert Burden, and former Coles Web analytics manager and now principal of Lens10, Naomi Le Get, shared how they brought together their digital analytics efforts to develop an enterprise-wide Web analytics strategy and approach at the supermarket giant.
Le Get was brought on as Web analytics manager in 2014 and tasked with developing and implementing a Web analytics solution for Coles Online. She said the emphasis was on understanding the online grocery shopping process, and improving basket sales through product marketing, brand actions and more targeted portal interactivity.
“Not only are you dealing with an average cart size of 60 items, but you can reorder your order several times before it is actually delivered,” she told attendees. “It’s a fantastic proposition for customers but that’s unbelievably challenging for analytics. We had an analytics tool in place but it wasn’t really dealing with this, and we needed much more grunt.”
In contrast, Burden sat within the Coles Financial Services business and was tasked with building its digital analytics optimisation efforts and better tie into the greater Coles business including Fly Buys and Coles retail.
The division was already using Adobe Analytics but struggling to achieve an end-to-end view of its customer. This was partly because it leverages a range of external partners to provide products to consumers including IAG, MetLife and GE Latitude.
“Data was paramount to driving the effectiveness of our marketing campaigns,” Burden said. “We had analytics already and were the first part of the business to roll it out, but it was broken.
“The challenge was there was no end-to-end data integration. We had four types of analytics across different websites and also running three different marketing clouds.”
Key to success was gaining a clear, unified view of the customer. To do this, Burden said he knew he needed to start from scratch, engage stakeholders as well as re-engage external partners. “We started thinking about the right plan to fix the hornet’s nest and have an accurate view of that end-to-end behaviour,” he said.
Le Get, meanwhile, was working on pulling together a couple of divisions together and ran a vendor procurement process for an enterprise-grade analytics solution. In parallel, a new grocery site was being developed that was full of rich data layers from day one.
“We figured a tool that could deal with complex processes in one part of the business could scale for the whole business,” she explained. “We realised we needed Adobe Analytics to deal with the scale, complexity and richness of data, then take it out to Coles as a whole.”
In Coles Financial Services, Burden was untangling analytics processes and tools and gaining clearer customer insights. He then implemented Adobe Target, a Website testing and personalisation tool, to leverage insights in order to drive better digital marketing activities.
Examples of activities Burden highlighted were A/B testing across onsite buttons, calls to action and product page content.
“The business was nervous about the tool – security had their issues – so to get across the line, we started a pilot program to prove firstly that it could drive lift, secondly, that it wouldn’t destroy the website, and thirdly, that we were going to get teams to work collaboratively,” he said. “That’s a challenge, as you need marketing, security, legal and so on all involved.”
Financial Services has since rolled out an ongoing optimisation program for all of its financial products, Burden said. To engage teams with the data insights being delivered, it also runs competition walls giving staff the opportunity to guess which test will win and resulting engagement uplift.
In Coles Online, Le Get said the team gained the ability to sequence people’s trolley builds, helping improve basket size as well as inform product choices made by customers. For example, by just putting bananas on its front page, Coles could nudge people towards healthier choices.
“If you add bananas first to your trolley, you will add an average of 18.6 fresh produce items in your cart. If your first item is milk, you’ll add an average of 13 fresh produce items. Order is really important,” she pointed out.
Getting cultural buy-in
Having the sorts of findings that could catch people’s interest, the next step was building a culture around that, Le Get continued. To do this, she tapped into the strong data DNA inside the business built over decades.
“The trading meetings at Coles are renowned,” she said. “Those were for physical store performance, however, not for online. But the grocery meetings meant we had a model we knew we could use and one understood by the wider business.”
Le Get highlighted two other “clever” steps taken upfront. The first was an enterprise-level contract with Adobe. The second was to design global templates from day one.
“That meant we had massive efficiency that could be accomplished just by a little bit of talking and networking within the building,” she said. “Every department had to fund and implement their own Adobe implementation. In this case, it was an advantage as it meant implementation was staggered. I could give them a template and take time to explain it, introduce them to consultants who had work on previous implementations and had used the template before in Coles.”
Another step that helped pull analytics across the enterprise was having weekly meetings to review key KPIs every month.
“Everyone was welcome, and we made sure the meeting was peppered these with insights that people couldn’t learn anywhere else but in that meeting,” Le Get said. “Within six months, we had senior managers from across the business coming to the meeting, and we were taking highlights of those meetings to other senior management meetings.”
A monthly forum was also set up between finance, Le Get and the Fly Buys team to start linking analytics efforts together, and see how each team could help the others.
The final jump to a shared enterprise function then occurred, with Le Get’s role becoming part of a broader digital shared group. Shortly afterwards, Coles Finance Services was absorbed into that group as well.
Burden said the first key learning from these experiences is that it’s important to get the foundations for analytics right first, including the tools and people that can drive it forward.
“Start small, surface that data early on, and make sure you align your key stakeholder’s KPIs,” he advised.
Optimisation is another important part of the analytics journey, Burden said. “It’s great to have data but what do you do with it? You need A/B testing to take action and put a revenue figure against these numbers,” he said.
As a final piece of advice, Burden said make data and analytics fun for others in the organisation.
“Get people involved in data, because if it sits in the tools, no one sees it,” he said. “And you don’t need an army to get started.”