Using google trends to gauge interest for the consoles, I noticed that search popularity correlates quite well with weekly sales ratios between the consoles.

PS4: 55

Switch: 35

XBox one: 23

3DS: 10

PS Vita: 2

Using the latest PS4 weekly sales estimate of 331,394 and multiplying by the search ratio you get 210,766 sales for the Nintendo Switch,  0.636 * 331,394.

The other results yield 138,582 for XBox one(77,502). 60,253 for 3DS(52,093) and 12,050(9,457) for PS Vita. It's not perfect but aside from the XBox One, they are within a decent degree of accuracy. Sales from vgchartz.com in parenthesis.

If I'm right, that would mean that the Switch is being under supplied by around 66,000 units a week.

https://trends.google.co.uk/trends/explore?q=%2Fm%2F0vphbfj,ps4,switch,3ds,ps%20vita

The Wii still has interest, does not mean that people will buy it. You need to take in consideration that there are other factors, notably, the interest of people that already have the console.

Be careful, you might have discovered the way VGC gets its numbers

Of course there are a myriad of other variables and factors, it's just an attempt to gauge what the Switch would sale if it could meet demand.

I've never heard of anyone using this methodology before. It certainly needs to be tested repeatedly before it has any credibility!

I don't think so, looking at media create trends of Switch sales recently, I'd more likely say the Switch is being under supplied by 66,000 per week in just Japan alone.

Neat idea, but it needs way more weeks/months of data to be tested and I suspect you'll find all kinds of inconsistencies around E3 and maybe holiday seasons.

It is an interesting method. I'm not sure you can use it for exact numbers since it will be very imprecise. It can still be useful as higher demand for a console will see people search for it more (such as research and buying the system online).

If you want to tell if it has any validity, run a regression analysis and get an R Squared. This will tell you if its useable (should have an R Squared above 70%) and how much a change in one will affect a change in the other.

The generic word "switch" doesn't only include Nintendo's new console. You can see that very clearly if you look up the Google Trend of 5 years:

So you can subtract a baseline of 18 - 20 from the 35 above since I doubt that people suddenly stopped searching for other switches. It is similar to "3ds", which includes Autodesk's popular "3D Studio" .