Big data. It’s the new oil. The north star. The saviour of business. But have we got ahead of ourselves? I’m worried that we have just given up on solving some of the most foundational retail metrics before we move onto the next shiny object.
Sure, machine learning, artificial intelligence and quantum computing are exciting, but let’s pause to reflect on the soldiers we have left behind.
Cross-channel marketing attribution
For years, marketers have been promised one total view of their customer journey across both online and offline channels. No one has cracked it.
The loudest fanfare came in 2016 with Google Analytics releasing Google Attribution 360. It promised marketers the ability “to measure and optimise marketing spend for all channels, online and off, at once“.
At $150,000 per year, it wasn’t cheap, but worth considering if Google had solved it. Unfortunately, Google Attribution 360 couldn’t live up to the hype and disappeared into the broader product suite.
Google isn’t the only one to fire and fail. There have been numerous startups that have attempted to solve the holistic marketing ROI attribution problem but never been able to gain market traction.
True cross-channel attribution is only going to get harder as marketing channels continue to diverge and shopping journeys get more complicated across online, in-store and marketplaces. With most marketing attribution questions still focused around the first click, last click or linear models, I fear this one is becoming too big to solve now.
Online to in-store sales attribution
There is still no agreed method to measure the influence online has on in-store sales. There seems to be a consensus that an omnichannel business model is more likely to be successful than a single channel one – but there are no certified metrics to prove this other than historical trends.
At the end of last year, Super Retail Group’s CEO Anthony Heraghty said their loyalty programs are suggesting that “only two per cent of customers don’t use a store at all”. While a reliable trend indicator, loyalty programs are naturally skewed towards more frequent and engaged customers.
Of course, there are other methods. I have seen a leading online ad network roll out hundreds of in-store beacons for free to prove the power of their ads beyond just online engagement. I have myself, out of pure frustration, placed researchers outside of physical stores asking customers if they researched online before coming in-store.
None of these methods is perfect. They are enough to prove the omnichannel effect, but we still don’t have reliable online to in-store sales attribution.
Customer lifetime value
Most brands still can’t pinpoint the customers who have the most influence over their sales. Sure, at the most basic level, customer lifetime value can be tracked by calculating total sales over time divided by the number of unique transactions, but this tracking does not usually extend into a customer’s profitability and their direct influence on other shoppers.
I don’t know about you but my ideal customer isn’t one that buys $2,000 worth of goods per year for $100 profit. Give me a customer that spends $500 per year at 50 per cent margin, rarely processes a return and tells five other people about their great experience.
This is my ideal customer. But which businesses have this kind of clarity on their best customers? It is very rare (and expensive). There’s certainly no off-the-shelf solution. With customers being ultra-aware of the data they are handing over, the locking down of social profiles and the rise of ad blockers, customer visibility isn’t getting any clearer any time soon.
The best solution doesn’t exist yet
These are just some examples of the fundamental retail metrics that remain unsolved despite the promises of a data-based future. Have we given up? I don’t think so. Is it getting any easier to solve? Unfortunately not.
If it is to be solved, we cannot rely on the ad networks to do it. It’s not in their interest to give complete transparency for hyper-targeting and zero-waste campaigns. Conversely, if retailers create complex internal algorithms that are not easily understood, they run the risk of self-driving their business into a wall. The ultimate solution is industry recognised solutions. They just don’t exist yet.
For now, it is up to retail to pick relevant metrics, keep it clear and follow the trend line. It means you will make some assumptions and not always have the full picture, but at least you know you won’t be alone. For now, anyway.