As a senior sales & marketing leader, I am drowning in analytics. Attribution analytics, Google analytics, sentiment analytics, pipeline vs. marketing spend, the list is endless (and probably analysable in itself).
Here’s my challenge: these “sources of the truth” tell me what happened yesterday. The leaders in my business don’t want to know about yesterday. They are focussed on what will happen tomorrow, next month, next quarter, next year.
Whilst analytics give me an accurate picture of sales and marketing accountability and performance in retrospect (critical at quarter end, when the teams’ bonuses are based on hitting targets), the success of any business operating in today’s economy is driven by futures. So where does that leave me?
Stuck in the middle with you
If like me, you have multiple streams of data about yesterday but your boss wants to know about tomorrow, you probably use analytics to deliver fact and insight for forecasting. That makes sense, it’s what we’ve always done, but let me ask you this: have you ever signed up to a financial product of any kind, and not seen see the following in the small print?
“Past performance may not be indicative of future results. Therefore, no current or prospective customer should assume that the future performance of any specific investment, strategy or product will be profitable or equal to corresponding indicated performance”
Everywhere you look, there are messages like the above, suggesting; “Just because it happened yesterday, don’t count in it happening again tomorrow.” Right or wrong, the markets expect organisations to forecast what will happen next
February, based upon analysis of what happened last
In the technology industry, this is even more flawed. A year is a lifetime (quite literally in the case of some products). So to forecast next February’s performance based upon last February’s is illogical. When you throw in curved-balls like Brexit and Trump and the unbalancing effect they might have on the market, it moves from illogical to impossible.
Forecasting vs. predictive analytics
What is the difference between forecasting and predictive analytics? Ask 10 people, you’ll get 10 different answers. Here’s my take:
is formulated and fixed. It supposes that drivers behind the demand for a product or service are intangible, so uses a statistical model that looks at the past to forecast the future; “we sold 10 things last February and we have demonstrated 20% year-on-year growth, so we forecast that 12 things will be sold next February”. It is sound logic, but not very agile. (Let’s not even discuss the issue that the “50% win probability” of sales person A can be entirely different from B, C, D, etc.)
are about people/buyers/shoppers and buying environments. They model the future by empathising with, and mapping the behaviour of, the people in the sale, as well as other factors in the sales cycle, not just by analysing the history of the item/service. “That’s witchcraft!” some will shout, but you would be surprised at the accuracy of the prediction. The more data you can use for modelling, the more accurate the prediction will become. This bodes well, as the tsunami of data coming our way (customer data, weather data, location data, etc.) increases exponentially by the hour, whereas the data from your results last February are quite limited by comparison and finite.
Which brings us back around to the starting point: I have too many analytics, my bosses want too many forecasts. What we actually need is a predictive analytics platform capable of soaking up the colossal data coming our way and automating the modelling process with “What if?” analyses, auto-feeding us insight and trends, not waiting to be asked for pre-defined reports.
This should be beyond “on demand”. It should be proactive – or would that represent artificial intelligence in the sales & marketing world?!