Post-Event Round-Up: “How Attribution Can Save Online Retail in 2021”

The impact of the pandemic has created mixed fortunes for retailers, generating both winners and losers. But a number of issues including increased commercial pressure, to changes in customer behaviour and the need to reach new customer segments is challenging retailers to be even more targeted and effective in their marketing.

Effective attribution is at the heart of this.

We were delighted to host an online webinar in partnership with Campaign to explore this in more detail.

The event theme was “How attribution can save online retail in 2021” and it featured a presentation and discussion between our CEO Chris Liversidge, our client Haroun Saleemi (Head of Ecommerce at QUIZ) and Suzanne Bidlake (Commercial Editor at Campaign).

So – what were the key takeaways from the event?

WATCH THE FULL VIDEO OF THE EVENT HERE

Retail winners and losers – and recent changes in customer behaviour

The retail sector is a fairly broad church, which means that the impact on the sector hasn’t been uniform.

Interestingly, research by KPMG and Ipsos the Retail Think Tank has highlighted the issue indicating that 60% of the UK market had grown in 2020 while 40% of it had declined. With food and home related items showing impressive growth as retailers benefited in a shift of spend from travel, leisure and hospitality sectors. However, spend in clothing and non-home related items has been impacted in the current environment.

This was echoed when Haroun pointed out that the current retail environment had been challenging for QUIZ and that lockdown had impacted the level of demand for the high street fashion retailer. But he also explained how they have been able to pivot the products they offer; increasing the focus on loungewear and casual products which sit away from their core occasion wear offerings.

Looking forward, the data is pointing towards a brighter second half of 2021. Retail spending is expected to bounce back– and Chris presented an overview of a number of recent trends in consumer behaviour including the fact that customers are increasingly:

• Hyper online
• Hyper social
• Hyper local
• Hyper connected

The key for retailers is to position themselves for the upturn in this adapted environment. In this model, there is a much stronger connectivity between the kind of expression of brands online and offline. And omnichannel approaches and measurement become critical. Something that the webinar audience have high on their agenda – with 67% strongly agreeing that the pandemic and changes to the retail landscape have made the need for accurate attribution more apparent for them.

A disconnect between omnichannel value and tech to support it

Attribution can be a difficult proposition for retailers and – as uncovered by research from Bright Pearl – there is a huge disconnect between having an omnichannel strategy and implementing it. With 91% of retailers saying they have the strategy in place or planned – and only 12% reporting they have the right technology in place to support it.

Haroun explained:

“I think one of the key issues really is complexity. We’ve been presented with a few attribution options in the past, and it just seems over complicated, difficult to understand what the model actually is. And if I can’t understand that, how can I communicate that across to the rest of the business?”

Chris also pointed out that the nature of the analysis can be problematic – “if you’re saying, suddenly, that a particular marketing channel is performing better or worse than you have been historically, that’s a challenging message”.

Underlying attribution data is poor – and siloed

As Chris presented, according to the IAB, 93% of all digital spend in the UK is allocated between Facebook and Google. The problem that retailers are facing is that these channels are separate in the way that they collect data and report it back to you.

So, using the real-world example below – and Facebook reporting at its most conservative – Facebook are attributing £450,000 in revenue from the campaign. While using Adobe or Google the number is closer to £20,000.

Haroun hit the nail on the head on the webinar when he said:

“If you were to kind of go by what Facebook tells you, then you wouldn’t stop spending. And if you were to go on last click attribution, then you wouldn’t spend at all.”

The reason for the disparity? There are a number of considerations here (and if you are interested in the detail check out the middle section of the webinar) but essentially the raw analytics data is actually extremely poorly put together. Rather than trying to identify the person behind the clicks, it focuses on pixel and cookie analysis associated with web visits.

This is the approach that the new breed of attribution solutions like our own attribution platform, Corvidae, takes – effectively using Machine Learning and AI techniques to rebuild the underlying data to provide a true and accurate picture of the impact on revenue.

And as Corvidae broadens the analysis out across channels, this provides retail marketers with the data to make different channel spending decisions – as the example for leading UK retailer, Tesco, shows.

Visit level attribution and driving ROI

So, as Chris pointed out, while the improvement in the channel attribution above is a huge step forward at a channel level,. What’s a real game-changer for retail brands is Visit Level Attribution which can be applied at the campaign level. That’s where you’re setting your media spends, associating your creative and setting bid rules, etc. – so that you can start to apply your strategy. And directly impact ROI.

And the results are worth considering. In the example below QUIZ were able to use Corvidae to reallocate media spend from poor performing campaigns to the best performing campaign in this set – taking 22% of media spend that was being wasted and generating an extra £1.4 million a year of revenue.

And the story is similar on Facebook. By taking 34% of wasted spend and reallocating it generates another £1.6 million on an annual basis.

As Haroun explained:

“Touching upon some of the other kind of issues that we faced, we’ve managed to reattribute about 60% of the revenue to other channels, versus GA. And so that’s immediately helped us to understand, some channels are performing better than we thought.

Before attribution, we would be doing a lot of A/B testing, and playing around with the different figures and different budgets, etc. But with attribution, we can remove all of that and be a bit more accurate and scientific. At the end of the day, we have to look at what is the impact on the bottom line – that net ROI figure.”

Introducing… Corvidae

Corvidae is our proprietary solution to the attribution problem and was the focus of the discussion around alternative attribution solutions at this week’s event.

Corvidae solves the issue with flawed data by reconstructing it from the ground-up, using machine-learning techniques to ‘see’ the individuals behind the clickstream data and get a clearer picture of their full conversion journey.

This, combined with cutting edge multi-channel modelling, results in attribution data retail brands can trust, helping them to make more informed decisions around their marketing spend.