24 Feb 16 Do you want data, or insight?

Have you ever wondered how analysts manage to give such precise numbers for future revenue or earnings?

Some random quotes to illustrate my point

• AInvest has trimmed Alpha Group earnings forecast for FY17 by 7.8%
• …lowered its blended ARPU assumption to 42.6 for FY17
• BInvest forecasted Beta’s earnings in FY16 to rise by 83%
• Revenue growth is expected to rise by 7.4%
• EBITDA margins will fall by 67 bps

All of these have very definite numbers, often precise to decimal points. But is it possible to be so accurate?

Actually, it’s impossible!

What matters is the trend or direction. Specific numbers in a particular quarter can deviate substantially due to a variety of reasons. Despite this, much effort is expended on preciseness – even when they know it’s futile.

A similar fetish for preciseness shows up in the corporate world. In the last two decades, we have worked with innumerable customers across industry verticals and geographies, and encountered many such fetishists.

We were once analyzing M&A behavior of a particular set of companies. The objective was to understand their appetite for M&A and what types of companies they might buy. As part of the analysis, we had to look at historical deals. Actual verified deal sizes were not available for all deals, especially the smaller ones. While our analysis delivered the insight needed for our client, one of the stakeholders was most upset that we couldn’t narrow deal sizes to the last million. We wasted a lot of time in getting more granular data – even though it was not material to our client’s strategy.

Another time we estimated financials for an unlisted company, by actually building up the P&L based on several data points (raw materials, power cost, etc.) that were available to us or acquired via field research. Naturally, there were assumptions and the final numbers were only estimates. But our client insisted that EBITDA margins of 18-20% were not good enough. He wanted the decimal point – even though the additional precision made no difference to his CEO’s decision.

We tend to forget that all estimates are just that, estimates. Any market size estimate, growth or earnings or GDP number is ultimately only an estimate. Even the government uses samples and extrapolates. Nobody, even the best analyst or research house can measure or predict these numbers precisely. Though some of them may be closer than others, the final number is still an estimate.

Yes, analysts can provide a long term trend, a range within which the numbers will fall, future growth direction, and other insights that are important for decision making. Yet, many people get stuck in the groove – and a lot of time, effort and money are wasted in trying to improve accuracy by that extra percentage point.

Obsessing about data and not the insight, is like missing the wood for the trees!