This article continues our series on mega-threats to the sales industry, inspired by the fascinating bestselling book by Nouriel Roubini entitled MegaThtreats: 10 Dangerous Trends that Imperil Our Future, and How to Survive Them.
The next mega-threat we’ll take up is the mishandling of data. This is a threat to any enterprise—or, for that matter, any organization or even a government.
Data could be generally defined as “facts or information.” It is used for evaluating situations, creating understanding, and many other functions. Data, in many forms, is what humans communicate to each other.
There are two basic ways data can be harmful: if it is mismanaged or if it is inaccurate.
Former Soviet Union president Mikhail Gorbachev once famously said, “Life punishes those who come too late.” If inaccurate data is utilized and acted upon, the threat is in the future, and rectifying it can definitely come too late.
Many consider artificial intelligence a general threat to society. The fact of the matter, though, is that it is only a threat if the data the AI system relies upon is inaccurate or misapplied. And unfortunately, the damage from use of inaccurate data won’t show up until the future.
An AI system can only act upon the data fed into it. If the data is incorrect, the result can be disastrous. If, for example, an AI system informs a battle group that they should attack an enemy area, but the enemy information the AI program utilizes for analysis is wrong, the attacking force could be headed for defeat.
When all is said and done, the first point of data evaluation is to accurately pinpoint “what happened?”
In the above example, we could say that the AI program is attempting to analyze “what happened?” Where did enemy defenses end up being placed? The AI system then uses the resulting analysis to advise the battle group where and when to attack.
In a medical example, a person has a heart attack. The doctor then requires accurate data on “what happened” with this heart attack to figure out how to treat the person. There are many other medical examples, and if data is accurate and extensive enough, a long-sought cure for a particular disease could even be discovered.
When it comes to sales, we want to know how many leads have been created, how many opportunities have been converted, how rapidly opportunities are moving through the pipeline, and how many opportunities have been closed.
We would also want to know if we lost any customers in a given sales period. In a SaaS business, how much did we gain compared to the number of customers that we lost?
It all begins with “What happened?”
Why Did It Happen?
In analyzing data, we not only need to know “what happened?” but also “why did it happen?” This is diagnostic.
If the data AI relies upon, in the above medical example, consists of extensive clinical studies covering every possible facet of a particular pathology, the results could be quite positive. We would probably achieve a better outcome than we did with, say, COVID. AI properly programmed might have prevented the many errors that occurred regarding COVID by individuals, companies and politicians. Hospitals incorrectly reported data, leading to vastly inflated figures for COVID-caused deaths. I personally know of one instance in which a person diagnosed with COVID died in a motorcycle accident, but his death was attributed to COVID. People knew this, but because the hospital received some kind of COVID-related bonus, it was reported that way.
Or, AI could be applied to the First World War—why did it happen? While there are all kinds of pat answers to that question, the total answer is quite complex. The writers of World War I history who lived at the time would already have formed some kind of bias so that data is already tainted.
In sales, we definitely want to know “why it happened” if a customer was lost. It might be because the customer lost business. We would want to know “why it happened” if insufficient leads were created—perhaps the website was down, or salespeople had attention elsewhere. With any situation, you cannot remedy an issue or change your strategy if you don’t know what caused the problem.
What Could Happen?
Now we arrive at the stage at which we could cause things to get better, which for me is the positive side. This would be: what could happen? If the data is correct, and we remove any possible interference to correct analysis, analyzing what could happen could lead us toward predictive analytics. What would happen if we do something? What would happen if we don’t do something?
For example, if AI were equipped with adequate data points, perhaps we might be able to more accurately forecast earthquakes in regions where they are a threat. Or forecast occurrences when it comes to other environmental issues such as climate change.
In sales, you need to create marketing and sales scenarios to discover “what could happen.” What could happen if you introduced a new product? What could happen if you launched a new campaign?
What Should We Do?
Lastly, adequate and accurate data should also lead us to the conclusions of what should we do? Such a question moves us from a reactive to a proactive stance concerning data.
For example, lack of food, especially for children, is a global problem. AI, provided with accurate and adequate data, could possibly lead us to better distribute what would otherwise be wasted food.
In sales, “what we should do” AI could provide some options based on available accurate data. For example, if you know how many leads you need to feed into the pipeline for making a particular quota, you might learn from AI that your weakness is not in the closing of opportunities, or the speed of opportunities through the pipeline, or even the conversion of leads to opportunities, but the fact you don’t have enough leads coming in.
Perhaps you don’t have enough meetings set, or you had enough meetings set but haven’t actually had them because prospects didn’t arrive. The solution might be to set up automatic meeting reminders to your prospects. With Pipeliner, these could be totally automated.
Answering the question of what should we do? leads us to prescriptive analytics.
Many CRM systems today, utilized by businesses for their sales teams, rely on AI to help predict sales. This will become increasingly true in the future. If the data used to guide and predict sales is incorrect, sales and the overall industry will suffer. For example, a company should have accurate data on the amount of time it usually takes for an opportunity to make it through the sales process to a close. If this data is not correct, every single opportunity could be mispredicted.
For sales, inaccurate and mismanaged data is a true mega-threat. Increasing numbers of sales solutions will be released as time goes by. No matter their efficiency, their power relies squarely upon data accuracy.
Straighten Out That Data
If we don’t become smarter, honest and more transparent regarding data, we will continue to be misled. If data is complex, we’ll constantly be heading in the wrong direction, running into the fire instead of running away from it, or jumping into an erupting volcano—all because of incorrect or mismanaged data.
I strongly believe that data accuracy is one of the areas that must be regulated when it comes to AI.
So the bottom line is: make sure that data is correct!