Right before we start, I want you to know that this isn’t a blog that you need to be a data scientist or an egg-head to read. I’m writing this for somebody like myself, the average salesperson, who is interested in reading about how maybe, just maybe, I can sell more than my competition.
If you are looking for some deep artificial intelligence article, this isn’t for you, I’m not a MIT lecturer.
Before we jump in and talk about a simple example of using machine learning to sell, a quick update on what the heck machine learning is.
Machine learning is software that can learn access, analyse data without being programmed and find patterns in that data.
The technology is being used in a wide range of industries for use cases including fraud prevention, predicting crop yields, preventing and mitigating natural disasters, predictive maintenance of enterprise assets, and improving supply chain efficiencies.
The concept of machine learning is not new; in fact, some machine learning algorithms have been around for decades, and machine learning algorithms are often used in predictive analytics.
Technical bit over, honest.
One of the changes that the internet and social media has made to the world is that people can (of course) and they do, research products and services like yours. One of the changes is that we ask questions of our networks as well as share information about ourselves.
In my book “Social Selling – Techniques to Influence Buyers and Changemakers” I quote some Google research which quotes the fact that text messaging had a revolutionary impact on society. People who grew up with text messaging, found that as it is such a quick, easy and cheap way of asking questions. They got used to asking questions of their network.
For example, “Our telephone system is rubbish and my boss has asked me to look for a replacement, any ideas?” We use Uberconferance for conference calls and collaborating, because I asked my network for advice.
Coupled with the fact many people are sharing their lives on social media. For example, if you get up to go to work and your car has a flat battery, the first thing you will do is go on social media and make a comment about it, before even fixing it.
Let’s call this “intent data”.
If people are asking for advice on telephone systems, conference call software or are saying their day has been ruined because they have a flat battery. There is probably an intent to buy? You might want to buy a new car battery after all.
You maybe thinking that this is all very good, but there must be a lot of noise out there. Totally. Which is where the machine learning comes in.
If like many people I speak to you are thinking that your customers are over 40 (please go with me on this and let me use that as a metaphor) and they don’t use the internet or social media. Totally understand. The Google research said that the people using social media in this way are between 28 and 35, they are trusted in organisations as they understand social and have business acumen. They have influence in an organisation but no authority.
The people “over 40” don’t do the searching they delegate it to this band of people. It is these people that are drawing up the short list. So selecting the short list as well as deselecting.
(Your clients may not be active on social, but the people who work for them (the next generation of business leaders, are.)
There is a tool by Microsoft as part of the Dynamics 365 range that you get it to listen for the terms and the criteria you set up. For example, “flat battery” and then it presents these to you in batches of 10.
Here is the clever bit, you can accept and reject the “leads” and the system learns based on the accessions and rejections. So if there is a “noise” out there on social, the system will get better and better.
Better still there is no rekeying, the “leads” will drop into Dynamics 365 CRM as MQLs (Marketing Qualified Leads).
Each salesperson can do this, so you can listen to something for your industry or product and do it and be self-sufficient in terms of your lead generation.
People who read this also read these: