The NPD Group recently reported alarmingly low U.S. retail sales numbers for January 2013. Gamasutra analyst Matt Matthews says there may be more to that story... As you may know, I deal primarily with hard numbers -- factual data from which I can draw conclusions. But sometimes, there's a bigger story behind the numbers, and I figure that it's important to share that with you, so you can try to draw your own conclusions. Almost as soon as Gamasutra published my column on January U.S. retail sales for Nintendo's Wii U, I got two messages from two different sources inside the industry both saying the same unbelievable thing: The Wii U probably sold significantly more than the 57,000 units reported by the NPD Group. How much more? The figure I was hearing was that the total number sold to consumers could go over 100,000 units, nearly twice the original figure. But it isn't so much that the NPD Group estimate was an error, I was told, but rather it didn't show the full picture. The figure reported by the NPD Group, the sources' story went on, included perhaps 100,000 units sold to consumers -- and 40,000 or more units returned to stores. The net, then, would yield the 57,000 units reported by the NPD Group. And the explanation for those tens of thousands of returns? The collapse of the secondary market, those resellers who had purchased Wii U systems in November and December 2012 hoping that popularity and a shortage of systems would yield a tidy profit through Christmas and into the new year. However, profiteers advertising Wii U systems on sites like Amazon, eBay and Craigslist saw their margins disappear and then chose to return their systems to retailers while their original receipts still permitted them to. So in the end, there probably still were a minimum of 57,000 units that ended up in consumers' homes in the U.S. this January. But it's nonetheless interesting to examine how these prospectors may affect the retail environment. Is such a thing even possible? Could the speculative market really have returned a volume of hardware that was 40-50 percent of the actual new sales in January? A quick check of VideoGamePriceCharts (which aggregates data from various resellers) shows that Wii U prices for new systems have been below retail since December 2012. And as of today there are numerous brand new Wii U Deluxe systems selling on eBay for well below the $350 retail price.

Digging deeper

Given that the original reports were based on NPD's data, I reached out to the company. Analyst Liam Callahan told me that he couldn't comment on whether the speculative market affected sales because he doesn't have access to that level of data. In order to protect their partners in the industry, the NPD Group maintains confidentiality of retailer-level data. Given my interactions with them before on issues that bordered on retailer specifics, I wasn't hopeful, but now I know there is absolutely zero probability that the NPD Group will ever talk about this issue -- even if someone there knows for sure. Then I contacted Michael Pachter, analyst for Wedbush Securities, to get his reaction. He said that consumer returns could count and affect the net figure reported by the NPD Group, but that the very idea sounded "like some Nintendo spin" to him. And, even if total new Wii U sales were 100,000 in January, that would put an estimate of 80,000 for February -- still not a healthy number. But what about software sales? The two original tipsters had independently pointed to software sales as evidence that the reported hardware figure was askew. That is, they hinted that total software unit sales were more indicative of a system selling 100,000 units in a month, not 57,000. I put that question to Pachter, and his response was pretty direct: Looking at the installed bases, all three of the previous console launches were much healthier in January software sales. And we're not talking just a couple of thousand units more -- we're talking about a factor of two or more in some cases. That's when I decided to call up a source I know who is familiar with a major retailer's sales at both a local and regional level. Without giving any indication of the tip I'd been passed, I asked the source directly about Wii U sales in January. The response: total new sales were about twice as big as the number of returns they had. In fact, it had gotten bad enough that some retailers were looking at ways to turn away people who appeared to be abusing the returns system. So, for example, if that retailer had sold 100 systems, then their returns in January were in the 40-50 unit range. That was eerily in line with what I'd heard from the original tips. So how do we settle this? The truth is, we probably won't ever know for sure what happened in January. Even the original sources admitted as much to me. But we can look forward to what's going to happen when the NPD Group reports Wii U sales for February. By that point most speculative resellers probably would have exited the market, and returns would have likely ceased to be a factor. If actual new sales of the Wii U had been 100,000 across five weeks in January, then after four weeks in February we could see sales as high as 80,000 units. If the speculative reseller return theory is wrong, however, and 57,000 is really reflective of the total number of new systems sold in January (a five-week retail month), then we could see a mere 46,000 units in February (a four-week retail month), if average weekly January sales are repeated in February. The NPD Group will report those February estimates in under two weeks. I'll be back then, and we can revisit this question.