Let's talk about account plans

Let’s talk about account plans.

I’ve seen my fair share of them, and they come in all different flavours and sizes.

I’ve even created many versions myself throughout the years.

And they all have one thing in common: As a tool used by a cross-functional team, all them fall short in terms of usability and delivering on intent. I haven’t heard one single person say: I really love working with my account plan template. It really helps me and my team reach ours and our customers objectives.

So why do we keep using them? And why do I keep providing my clients with what I think is a good account template I recommend them to use?

Well, I guess it boils down to one thing: The CRM systems of the world are not designed to support team-based key account management work.

Now, that view is purely on me. Let’s look at what the research says about account planning.

[INSERT: Gartner, 2024, on why plans get abandoned: account managers find them “burdensome”, “performance goals don’t encourage plan maintenance and updates”, and lack of “training, coaching, and ongoing support reduces acceptance and increases abandonment”]

[INSERT: Gartner: 66% of key account managers agree the amount of internal coordination limits their ability to sufficiently strategise for account growth.]

A couple of weeks ago I joined a team during their monthly account review.

The week before the manager sent me a copy of the updated plan that was up for review and since they used one of my account plans so there were no surprises there.

As I looked through the plan I realized something: I skipped 95% of the content and went straight for what I think is the most important part of any KAM-plan: What is the win-win-win opportunity we believe exist at this account and how do we pursue this opportunity together with the relevant account stakeholders?

For me this is the only thing that matters. This is the big WHY we have a whole cross-functional team in place. Because if the team is successful in this pursuit, there’s a win for patient, the account and their company waiting.

During the review this was also what we discussed and the output generated included a specific action plan designed to push the identified opportunity forward.

In all aspects, the account review I just participated in was doing its job getting team one step closer towards the goal. But there was something in the back of my mind I couldn’t let go of…

A couple of days later I read a post that asked: Do you know why Anthropics AI is called Claude? Since I didn’t know I kept reading. It turned out that Anthropic, most likely since they haven’t confirmed it, baptized their AI after Claude Shannon, a brilliant scientist behind the theory of information.

And then I went down a rabbit hole…there’s a lot to say about this theory, and quite frankly I didn’t understand it all because it can become quite complicated, but for the purpose of what I’m trying to do here, here’s a high level summary:

• Information is not about content. It is about choice from alternatives.
• If I can only send you one possible message, and you already know that, then sending it tells you nothing. The number of possibilities = 1.
• If I could have sent any one of a million messages, but I send one, that tells you a great deal, because you had a million alternatives to start from giving you a high level of uncertainty. After receiving the message your uncertainty is reduced to zero.
• To conclude: the information content of a message is the amount of uncertainty it removes. And the amount of uncertainty it removes depends entirely on what you did not know beforehand.

Shannon had a name for this uncertainty, which can be quantified: entropy

High level of entropy = you have no idea what the content of the message is.

Low level of entropy = you know beforehand what the content of the message is.

So why did Anthropic name their AI after Claude Shannon?

In 1948 he published a paper which concluded that: The redundancy of ordinary English, not considering statistical structure over greater distances than about eight letters, is roughly 50%. This means that when we write English half of what we write is determined by the structure of the language and half is chosen freely.

And that is one of the reasons why AI and LLMs work.

Why do I tell you all this? Well, as I mentioned, I had something gnarling in the back of my head after the account review, and I suddenly realized it was because of Claude Shannons work on information. And it raised a number of questions.

Why do most of us hate writing KAM-plans?

Number 1: The plan is not for us. We already know the stuff we are putting in there.

Number 2: We know it’s for our managers; and they just want to make sure we are working and have something to show if someone would ask.

Back to Shannon: Based on this, we can say that a KAM-plan is used to reduce the uncertainty for everyone else except the author. And that once a KAM-plan has been distributed and consumed, it becomes practically useless as a carrier of information.

There are exceptions, but for the sake of argument, let’s leave them out of the picture right now.

In one of his articles, Shannon stated, rephrased by me: Information does not exist without a specified set of alternatives which means that before you can measure what a message is worth, you have to write down what else it could have been.

What are the consequences if this? The most valuable output of a knowledge base like a KAM-plan is a map of ignorance, not a set of known facts.

As I came this far penetrating the depths of information theory I stopped. Because here was something I actually advocated for when training cross-functional teams.

Begin with what you know and what you think you know.

Then you start working on the most important part: What is it you don’t know that would make a big difference in your relationship with the account and its stakeholders?

Capturing this in a KAM-plan would clearly be valuable to the cross-functional team since this represents high level of entropy and uncertainty. Now, to be fair, Shannon only counted how much uncertainty a piece of information removed, he never cared about weather that reduction generated value or not.

So, based on all this, there’s clearly utility in having a KAM-plan but what the actual utility is depends on the reader. A manager is looking for certain things. A consultant like myself for a different set of information.

This reasoning leads me to the following conclusion:

Yes, there’s value in a KAM-plan, but the value is not what we think it is.

If it was possible, we should use an account information foundation we can extract what we need depending who is interested. A set of unique slices of plan that carries value for the recipient.

Coming this far I began working with our present day Claude. I asked him to help me build a Shannon account plan compressor. Here’s the idea: by applying the principles behind Shannons information theorem on a key account plan, we will build an instant version of that plan that maximizes the utility for different users. A plan for the cross-functional team will look very different from the one a manger is reviewing.

Eventually, I had a prototype that when applied to a real account plan generated exactly what I wanted. In fact, it turned out soo good that I’m convinced anyone seing this will drop tradtional account planning in a hear beat. And no one would be happier than the teams that are supposed to write and update them as well as the managers who finally will see and understand what is really happening at the account.

To you dear reader

If you, like most people, want something better than the traditional KAM-plans, something that really drives business forward providing value for patients, healhtcare systems and your company, send me a DM, I’m looking for ONE partner to develop this into an operational tool that will change how we do account planning forever.