The Trouble with Traffic
by Mark Ryski
Whether same-store sales are up or down, analysts want to know what drove results. If you've listened in on an earnings call for a major retailer lately, undoubtedly you've heard the question: was it ticket or traffic?
It seems Wall Street analysts who poke and prod retail executives during the sometimes contentious Q&A sessions have distilled the “what drove same-store sales” question down to these two variables – either more people coming into the store and/or selling more stuff to the buyers.
The “ticket or traffic” question is certainly relevant, so when analysts ask, retail executives are compelled to answer. Answering the "ticket" part is simple enough – any POS system can produce the answer to whether average ticket values have increased or not. But what about traffic?
While virtually every retailer provides an answer to the “was traffic up or down” question, here's the rub: most retailers don't actually measure traffic in their stores. How’s that possible? Simple: traffic, it seems, has more than one meaning.
When a retailer is asked if traffic is up or down, there’s a very good chance that the answer provided actually refers to the chain’s “transaction count” or what is sometimes ambiguously referred to as “customer count.” No one seems to probe on this so, by default, transaction count has become an acceptable proxy for store traffic count. But there’s another rub: transaction count is not the same as traffic count.
Transaction Counts vs. Traffic Counts – Hits vs. At-Bats
To say that transaction count represents a reliable proxy for store traffic is analogous to saying that hits are a reliable proxy for at-bats in baseball. Yes, the two stats are related, but they are not proxies – not even close.
If baseball statisticians only tracked hits, without considering at-bats and batting average, how much less would we understand about the greatness of players like Ty Cobb or Babe Ruth? A lot less. The same is true for retailers. Transaction counts (hits) may be up, but knowing if it was a result of an increase in store traffic (at-bats), or that the retailer was more effective at converting the store traffic it got is an important distinction. This is not a subtle point. Here’s why.
Why Store Traffic Matters
Store traffic is a measure of all the people who visit the store, including buyers and non-buyers. Traffic is a leading indicator that tells us something about a chain’s sales opportunity – more traffic, more opportunity. If traffic is trending up, this is clearly a positive sign. It suggests that the brand is in favor and opportunities abound. The converse is also true. If store traffic is waning, this is disconcerting and it could indicate that the banner is falling out of favor. The number of sales opportunities is decreasing. The problem with relying on transaction counts as a proxy for traffic is that they could be going up regardless of whether actual store traffic is going up or down. To understand this apparent paradox, you need to consider the retailers’ batting average.
Conversion Rate – Retail Batting Average
As mentioned, store traffic count defines the sales opportunity and is analogous to at-bats. Transaction count represents buyers only and is analogous to hits. So, then, a retailer’s batting average, or conversion rate, is calculated by dividing the transaction count by the store traffic count – just like in calculating batting average.
Store traffic and conversion rates tend to be inversely related. When store traffic falls, associates are able to deliver a higher level of service, check-out lines are shorter, and generally it’s easier to buy. The transaction count often goes up, despite the fact that there is actually less traffic in the store. In this case store traffic didn’t increase, but if the retailer only has transaction counts to rely upon, then he reports “traffic is up”. But it’s not and all parties – the retailers and the inquisitive analysts – seem to tolerate the ambiguity.
Don’t Ask, Don’t Tell
One Wall Streeter told me that you can’t ask a retailer about traffic counts if they don’t track traffic in their stores. True, but you also can’t have two definitions for this basic metric either. If you want to ask about transaction counts then ask for transaction counts; if you want to ask about store traffic, then ask for store traffic. This shouldn’t be open to interpretation.
There is a simple way to inject clarity into what has become a convoluted question. Instead of asking retailers if it was “ticket or traffic” that drove results, analysts should ask if it was “ticket, traffic or conversion”. While most retailers don’t track store traffic and so won’t be able to answer, at least it will be clear that they don’t and you will know they mean transaction count – which on its own tells us little about what drove results. As for the retailers who do track store traffic and measure conversion rates, you will have a much deeper insight into what actually drove sales results.
Retailers, and Wall Street, need to take a page out of the baseball playbook.
Mark Ryski is author of Conversion: The Last Great Retail Metric and When Retail Customers Count and founder of HeadCount Corporation. For more information, please visit www.headcount.com.