Konnichiwa, I interviewed a former senior analyst at NPD here:
http://vgchartz.com/news/news.php?id=740
Here is the interesting part:
TheSource (3:03:18 PM): Sounds reasonable. Can you speak about NPD's procedures then and now? Have they changed? I'm aware that NPD adds and subtracts retailers on occasion, but I'm curious as to the market share changes in NPD coverage over the years.
TheMarketSage (3:06:57 PM): Sure. Essentially the core methodology hasn't really changed. NPD recruits retailers to provide them with SKU level data on a monthly or weekly basis. NPD then maps that data to the appropriate products and aggregates the sales totals up so that customers can see sales and the product and brand level.
NPD also applies an adjustment to the figures based on their estimate of the coverage of their retail panel. The panel was better (i.e. more representative) than it is now because it used to include Walmart. Plus, the Internet, and its fragmented distribution of sales, has made it tough to accurately assess the true size of the market. Thus, while NPD once had perhaps 85% of the market represented in its panel, that figure is now closer to 60-65% depending on the category. The videogames category is probably around there.
By the way -- NPD doesn't really subtract retailers by its own decision. That happens because retailers leave their panel. It is not usually in NPD's interest to cut someone out of the panel. That just lowers their coverage.
TheSource (3:09:28 PM): Well, I was thinking of the recent report of Toys R US leaving. What you said makes me think of a couple things... First of all, why does some data come in weekly, while some data comes in monthly? Second, how much is data really worth if Walmart and others occasionally opt out?
TheMarketSage (3:12:04 PM): Good questions. NPD would prefer all data to come in weekly. However, some retailers just aren't equipped or simply flat out refuse to provide it that way. Ultimately it is the retailers’ decision though.
As to the question of the NPD data without Walmart? Many would argue that it just isn't that valid, but when left with no alternative (until you guys came along), it is better than nothing. The validity is a hard pill to swallow for many though; particularly given the steep pricing customers are forced to accept.
TheSource (3:13:44 PM): How steep is this pricing? Who are the main customers? It would seem if there are only a few customers, then the pricing has to be steep to make it worthwhile to gather the data in the first place.
TheMarketSage (3:17:22 PM): It's a real balancing act. Certainly they need to charge enough to attract the perfect amount of customers to more than cover the costs of producing the data. Charge too much and the customers leave and you are short. Charge too little and you can leave money on the table. That said - I'm not exactly sure what they charge for their data (though I could find out easily). I know that some big clients like Sony were paying $1.5 million each year, but that was for a bunch of categories. My guess is that a yearly subscription to their monthly gaming data would be in the $50k range. However, they do have quite a few customers. MS, Sony, and Nintendo are all at the top of that list. It is likely that EA, Activision, THQ and the other big publishers also subscribe. Many of them just don't see an alternative. That is why you guys are so important. However, I guarantee you they won't ditch NPD in favor of you until they are comfortable with your methods.
TheSource (3:19:04 PM): Well, then let me ask you what brought you to Vgchartz in the first place? How do you use it? Currently, we have about 2-3% hardware coverage from a number of small retail samples that arrive weekly.
People are difficult to govern because they have too much knowledge.
When there are more laws, there are more criminals.
- Lao Tzu
