ROAS or Reach? A Case for Both

The media we use to drive eCommerce sales is a broad church. It ranges from Google shopping, Facebook collaborative ads to sponsored networks like Criteo and online media direct from retailers. The thing is, depending on where these media owners traditionally start in the funnel, KPIs are significantly different. 

Why is brand media like Facebook targeted on Reach but eRetail media from Tesco targeted on ROAS? This has something to do with the fact that the former has sat with marketing and media owners and the latter has sat with the sales teams and retailers. The former group are building brands and the latter are driving commercial success.  However, eCommerce doesn’t sit exclusively in either the brand building or commercial task - it straddles the two. Particularly now Covid is ensuring that an increasing part of the brand experience, as well as sales volume, is happening in ecommerce. This is why I believe Reach and ROAS must be used together to measure efficacy across all media intended to drive ecommerce success. 

Brand media is fantastic in helping us build a brand but leaves us wanting when we need to prove the commercial return. I have seen a lot of media get onto a brand plan to meet a conversion brief without ROAS as a KPI. However, I recently listened to Google speak very eloquently about the logic of thinking beyond ROAS. Their belief is that customer experience and traffic are the key drivers of retail success and if you get this right then sales look after themselves.  The extension of this that by using ROAS you are limiting traffic and limiting your sales potential.  The traffic and experience point is very much in line with thinking for those brands who have been considering a version 2.0 to their DTC site. After setting up very quickly and prioritising back end logistics, the next steps are to improve customer experience and traffic to drive better sales results.  

This argument is valid and as old as marketing itself, but for me any media designed to drive ecommerce success, be it brand or eRetail, must be accountable for a commercial sales result. Rhys Jones of Retail Alchemy, a specialist in this area, puts it very well “Ultimately, whilst 'longer term' metrics such as awareness or engagement are important, ultimately improvements in such metrics should translate into revenue further down the line meaning that evaluation of all forms of marketing should include a measure of ROAS.”

Amazon Advertising have done a good job. Fair do’s, they have an understanding of multiple stakeholders needs and have clear KPIs across Awareness, Engagement and Conversion.  This is not unlike the tale of Amazon in general - they are generally good at getting “it” before anyone else does. One tendency I have seen though is that they separate Reach as a KPI for upper funnel and ROAS as purely a KPI of their lower funnel media. Quite logical, but personally, if my goal is to drive sales on Amazon, I want ROAS to include both areas of spend, or it will boost the return artificially. 

The Citrus system that has recently gone live for Sainsbury’s replicates this Amazon Approach.  It has well-balanced and wide-ranging measures in Reach, Impressions and ROAS and more besides. It’s encouraging that Omni-channel retailers are matching the Amazon benchmark and this places them in good stead for the future. After all, for many manufacturers the vast majority of their ecommerce sales don’t sit on Amazon.

Criteo have long been an early leader in data provision with a focused approach on ROAS. This is a good thing, because many retail media owners don’t provide ROAS information regularly. However, in work I have seen to date their offer on the reach of their media is not as detailed, but with a new SSP recently announced I am sure this will develop much more in 2021.

The only thing that really concerns me about some of these advanced eRetail media models is that often you are not taking the effect of price mechanics away from the media returns. There is a great deal of investment behind this (retailers determine prices) and media needs to be conscious of not taking credit for the lever of price promotion. This is something Rhys Jones has clear data points on,  “Our research shows that as much as a quarter of brand sales can be driven by promotions alone, meaning that if such activity is not stripped out, media ROAS can potentially be overstated to a similar degree”. 

It is more than price promotion that complicates understanding the efficacy of media. Often I have been in meetings and we have been scratching our heads at a sales result. We had the right media plan; the offer was in place - but no uplift. There are just so many different factors in play. Salim Bachatene of Data Impact puts this very well “In order to fully analyse the impact of a campaign it is really very important to have in hand not only the information of the campaign but also other elements connected to the campaign such as out-of-stock, promotions and competitors actions”. The inference is that other levers beyond your control affect the impact of media on ecommerce sales, so don’t always expect immediate clear answers in your data.

Once you dive into the detail of retail owned media, you will find retailers have very different definitions of metrics as well as varying levels of capability to provide this data. One well-known retailer attributes ROAS based on revenues generated by whoever saw an ad whether they were already intending to do so or not – so not reflecting incrementality. Citrus just looks at revenues driven by a particular SKU that is included in a media buy but it is common to see media owners play the halo effect of attributing sales driven across a range of SKUs across the brand. I have worked in this area for some time and I could not be sure of all the different ways metrics are calculated. So for those businesses that are considering up weighting spend in this area, it literally pays to be clear on definitions or risk making an unwise investment.

Navigating the different metrics across media is important for us all to bear in mind when we are driving sales for ecommerce. It is not simple and it takes time to understand fully. So what does this all mean to the planners of media and client side marketers trying to drive ecommerce sales?

  • Recognise both Reach and ROAS as measures of media efficacy on all campaigns

  • Be demanding more from brand media owners who are selling commerce solutions to prove and deliver harder commercial results. 

  • Be clear on how each media owner defines ROAS

Enjoy the rigour and discovery of the journey.