A couple of years ago, Heavy Reading carried out a multi-client survey for half a dozen clients on the subject of public access small cells and small cell backhaul. We proposed the survey to clients because of the almighty swell of hype that had built up around the idea of extending the principle of the closed user group femtocell out from the home and the enterprise and into the public domain where these small cells could be used by any of an operator's mobile subscribers. Who, for example, could forget Picochip 's epic assertion in October 2011 that Chicago would need 84,500 small cells by 2015 "to deliver truly high speed LTE" and that this would be "in addition to residential femtocells and WiFi"?
Our clients asked us to carry out an online survey of 100 qualified mobile operators worldwide, combine that with a series of more detailed face-to-face and telephone interviews with folks engaged in RAN and backhaul network planning in the mobile operators, and reach our conclusions as to the direction of the public access small cell market, the likely rate of adoption, and the role of the backhaul network in this evolution.
The findings that we shared with clients in April 2012 proved to be right on the money. Whilst the theoretical logic of leveraging public access small cells for capacity fill-in was clearly compelling, operators flagged a number of major practical operational and other business case constraints such as site acquisition, RF performance at street level, installation issues, cost points, and coordination with the macro layer. This made it clear that nothing much by way of volume deployments was going to happen for the next two or three years.
The online survey results also yielded two other crucial findings. The operators told us that they were heavily focused on re-using existing backhaul infrastructure and decidedly lukewarm about investing in dedicated new backhaul for small cells -- be it fiber, copper, or radio. But as any cellular network planner will tell you, a public access small cell can't go in roughly the right location. For the business case to have the best chance of kicking in, the small cell needs to go where it needs to go. And that required a far greater willingness on the part of the operators to invest in taking the backhaul to the small cell, rather than count on being able to take the small cell to the backhaul.
The survey also yielded a marked disparity between the outlooks of operator respondents that classified themselves as RAN engineers and the responses of the survey sample as a whole. The survey showed RAN experts to be much more likely to cite site acquisition and other operational issues as barriers to public access small cell deployment and also markedly more concerned about interference with (and coordination with) the macro layer.
In other words, the guys who were going to be on the front line rolling out these devices according to a viable business model were easily the most skeptical about the deployment timeframes. Yes, yes, we recognized that small cells are more threatening to an RF engineer's skillsets than those of other personnel in the operator. But we only attributed a portion of their negative sentiment to that kind of narrow self-interest. And amongst the reasons for that is that small cells don't dilute RF planning skills to the extent that some people think (or would like to think). We emerged from the study, blinking into the daylight, in April 2012, with a forecast that there would be no more than half a million public access small cells in live service by the end of 2015, a definition we later refined to half a million public access small cells in live service requiring new backhaul by the end of 2015.
Being very conservative compared with many other analyst forecasts at the time, our forecast was about as popular as the proverbial turd in a punchbowl with some clients. We just didn't "get it," according to one, who didn't sponsor our market research (or indeed anyone else's). Another pressed us pretty much every quarter for something more favorable, becoming ever more pink with every passing quarter. An unimpressed and exasperated competitor from another analyst firm actually demanded to know how we could possibly "expect to help grow a market by being so 'down' on it."
So why did Heavy Reading call this market correctly (at least so far)? Well for one thing, on the basis that timing is everything in business -- particularly in the hi-tech sector -- we see it as our goal to support telecom sector clients in making the right investment decisions at the right time. And on that score, we served our clients exceptionally well with this study. Three of the six clients that sponsored our study were executives that were under varying degrees of pressure from investors or peers in the management team (or both) to "do something big" in small cells.
They didn't and they still haven't. Instead they've continued to target scarce R&D resources at where the operators really are spending, are now watching the public access small cell pioneers take the arrows in their backs, and are poised to capitalize once the market does take off (which in one shape or another it will). Did these clients hold back entirely because of Heavy Reading's study? Of course not. But was the study between somewhat and very valuable in helping clients argue the case internally and externally that the operators weren't ready to spend big yet? Yes it was.
It would be nice to think that what made Heavy Reading call this right in this case was the overwhelming superiority of our analysts. Nice but not true -- well, not entirely anyway. And (whisper it…) there has even been the odd occasion when we too have been overly aggressive or overly conservative with our forecasts.
But one of the things that gives Heavy Reading its sustainable differentiation -- what we were able to leverage to tremendous effect in this public access small cell study -- is our access to Light Reading's database of registered readers. When Heavy Reading sends out a survey invitation it goes out to thousands of Light Reading's registered readers within service providers worldwide. Once we've filtered out all the survey responses from people who claim to work for Vodafone but actually have an @largeequipmentvendor.com email address, we end up with a survey sample of anywhere from 70-150 fully qualified operator respondents from around the world for any given Heavy Reading survey.
That's a pretty decent sample. That in itself doesn't mean we inevitably go on from there to reach the right conclusions and make accurate forecasts. But gauging market demand on the basis of the disposition of a large sample of buyers rather than the aspirations of the sellers does give you a hell of a head start. A happy new year to all Heavy Reading's clients.
— Patrick Donegan, Senior Analyst, Heavy Reading