Demystifying the Numbers behind’s AppExchange
Stephen Cummins

Hi Harry

Thank you very much for the positive feedback and the great question. The goal of the article was to “demystify the AppExchange and it’s numbers”. The reason its mysterious is that it’s almost like a detective story working it out as the sources are varied — declared financials, slideshares in the public domain, public quotes by senior Salesforce people, Dreamforce presentations/videos, the live data on the AppExchange itself, historic data (for trends) etc. And at the end of the article I quoted Descartes who said “”Perfect numbers like perfect men are very rare.” The goal was to finally give people a feeling for the scale of things and the trends. For example almost everyone thought that the number of apps being added to the marketplace was accelerating. In fact the rate of addition was steady for a few years, but the average revenue from apps was accelerating considerably. The vast majority of the data was published or calculated and extrapolated from published data. Where an assumption is made, I point that out. The article was written 14 months ago. Almost all of the stats were derived from publicly available financial data (Salesforce being a publicly traded company on the NYSE). Some of the data comes from public statements by Salesforce personnel. That has since been redesigned, but for example you can see as I type that there have been 4,420,699 installs from the debut of the AppExchange until the moment I write this .A lot of the other stats are calculations based on financial data. For example salesforce breaks out where it gets its revenue from, we know how many installs there are and we know what percentage of the apps are free etc etc. Sometimes we can’t get things exact, but you can make smart estimates that will certainly be close to the right figure. The idea of the article is to give people an idea of the scale, trajectory and evolution of the AppExchange. Not to be precise — although much of the data happens to be precise. For some reason nobody had ever done this before and I felt people needed to get an idea of the scale of things. A lot of the data was from a live dashboard-like piece on the homepage of the AppExchange. Here’s an example of something that is easy to extrapolate: “The rate of app downloads is not accelerating. However the average value of those downloads is accelerating fast. By the end of September 2013 app downloads had hit 2 million. It took another 19 months until the close of April 2015 to hit 3 million downloads. And it has taken nearly 9 months to hit 3.5 million. Hence the numbers are stable, and from that we can extrapolate that current downloads are 666,000/annum.” The one big assumption I made was to guess what percentage of downloads were free apps (but I’m transparent about that in the article). “$1.5B annual revenues / (660K downloads x 56% paid apps) =approx $4,000. However if we make a reasonable assumption that that freebies are downloaded 4X as much as paid apps, that would push the download ratio from 50 : 50 to 80 : 20 and that leaves the average AppExchange transaction value at $10,ooo. This assumption is difficult to get correct (even approximately) and is by far the biggest leap in this article, but it’s not unreasonable. Just consider how often the Salesforce 1 mobile app, all the various packaged dashboards and other cool stuff from Salesforce labs are downloaded. All free.” My career in Salesforce spanned over a decade and I run and scale the 2 fastest growing LinkedIn SaaS groups in the world ( and Salesforce Evolution) and if I don’t attend the presentations for investors when in Dreamforce, I watch the videos or look at Slideshares from Dreamforce (another source of data)— so I obviously have a good feel for the numbers. When projecting into the future I drew upon public statements by senior Salesforce staff e.g. “Neeracha Taychakhoonavudh (Salesforce’s SVP of Partner Programs) says Salesforce will boost AppExchange revenues 5X over the next 5 years. Let’s assume they’ll take the same average cut. The data tells us that app download rates are neither increasing nor decreasing (yes it does!), but we know revenues are accelerating at a clip. If Taychakhoonavudh’s estimate is correct and all other trends and assumptions hold, then the data tells us that the value of the average transaction to a vendor will be $8,500 x 5 = $42,500.” And of course I quoted from IDC’s detailed projections on the Salesforce economy.

Regarding the assumption in one of the calculations, there’s good reason to believe I might be in the ballpark. I make predictions all the time about SaaS, and about Salesforce in particular. So far every prediction has been accurate. That won’t last forever (I’ll inevitably get stuff wrong), but there’s reason to believe that the strike-rate will probably remain high. You can see a lot of the predictions in these articles:

I really hope this helps to further clarify things.

All the Best, Stephen Cummins

Show your support

Clapping shows how much you appreciated Stephen Cummins’s story.