The positioning problem when your category does not exist yet

One of the biggest marketing challenges I have faced recently is entering an extremely busy market with a product that does not fit neatly into any of the existing categories.

Sales software is crowded. There are CRMs, conversation intelligence platforms, meeting recorders, sales engagement tools, forecasting tools, enablement platforms, sales automation software and thousands of newer AI sales tools. Some are established companies with strong brands and existing distribution. Others are startups backed by significant investment and spending aggressively to create visibility.

Normally, entering a crowded market gives you a fairly clear positioning problem. Who are the alternatives? Where are they weak? What makes you different? Which buyers care about that difference?

What I have found more difficult is a different version of the same problem: what happens when buyers do not really have a category for what you are selling yet?

That is something we are dealing with at Optivian.

We call Ollie an AI sales co-worker, and we increasingly think about the broader space as AI deal execution.

By AI deal execution, I mean AI that understands the context of active sales opportunities across CRM data, meetings, emails and other sources, and then helps sellers take the actions needed to progress those deals. It is different from AI BDR tools focused on generating new pipeline, and broader than an AI sales assistant or co-pilot that mainly waits for the seller to ask it for something.

The problem is that all of those distinctions make much more sense after you understand the product than before.

And positioning has to work before the explanation.

Buyers start with what they already know

When marketers work on positioning, we spend a lot of time thinking about what we want people to understand. The buyer starts somewhere else.

They hear a few words, combine them with what they already know, and form an assumption.

If we say AI sales co-worker, someone may first hear AI + sales, which today can very easily become AI BDR.

That is understandable. AI prospecting and outbound agents have received huge amounts of attention, and companies like Artisan have helped make that interpretation of “AI sales” highly visible.

But Ollie works primarily once there is already a deal in the pipeline. It builds context around the opportunity, keeps that context current, identifies gaps and risks, helps prepare the work needed to progress the deal and supports the seller in executing it.

That distinction is easy to explain once someone gives you a few minutes. The marketing problem is what happens before those few minutes.

This has made me think about positioning slightly differently.

You are not only competing against other products. You are competing against the category the buyer puts you into before they understand you.

The obvious response is to use comparisons, and there are plenty available. Parts of what we do overlap with conversation intelligence, because meetings are an important source of deal context. Other parts overlap with deal intelligence, deal health, forecasting and sales automation. Ollie can also help with MEDDPICC, follow-ups, mutual action plans, business cases and CRM updates.

You can describe parts of the experience using familiar terms like AI sales assistant or AI sales co-pilot. Each of those gives the buyer something they already understand.

The problem is that every comparison also narrows the product.

If we talk too much about conversation intelligence, we can become another Gong alternative in someone’s head. If we talk too much about meeting capture, we become a meeting recorder. If we lead with business cases, we become a business case tool. If we call it an AI sales assistant, it can sound like something that waits for prompts rather than actively helping move a deal forward.

All of those descriptions are partially correct. None describes the product particularly well.

That creates a positioning tension I had not appreciated enough before working on this:

If you avoid existing categories completely, people struggle to understand what you do. If you borrow too heavily from them, they can define you too narrowly.

The job is not simply to find the closest competitor. It is to decide which existing mental models are useful enough to borrow, and where you need to deliberately break away from them.

Positioning and demand generation do not always want the same thing

This gets even more interesting once you move from positioning into actual GTM.

You can decide that you are creating a new category. You can give it a name, define it clearly and start educating the market.

That does not mean buyers suddenly start looking for it.

Very few people wake up and search for something they do not yet know exists. They search for problems they already recognise.

A sales leader might search for ways to improve CRM adoption. Another might be looking at deal health. Someone else might want better MEDDPICC visibility, more consistent follow-up or a better way to create business cases. They might be researching sales automation software without yet knowing that what they actually need is something that understands the whole deal rather than automating one isolated task.

Those are useful demand entry points for us.

From an SEO, AEO and broader demand generation perspective, ignoring those existing topics because we want to own a broader category would make little sense.

But there is a catch: the language that helps people discover you is not necessarily the language you eventually want them to use to describe you.

This means we effectively have to operate on two layers.

At the demand layer, we can meet buyers around existing problems and established categories. At the positioning layer, we have to show how those problems fit into something larger.

For Optivian, CRM updates can be an entry point. Deal health can be an entry point. Business cases can be an entry point. Conversation intelligence can provide useful competitive context. AI sales assistants and co-pilots provide another familiar reference point.

But none of those should become the boundary around the product.

The broader idea is deal execution.

A deal is not a collection of disconnected tasks. The meeting affects the follow-up. The follow-up affects the next step. The business case depends on what the buyer has said. MEDDPICC gaps influence what the seller should find out next. CRM information should reflect everything happening around the opportunity.

If AI has enough persistent context about the deal, it can work across those tasks rather than solving each one as an isolated feature.

That is the shift we actually need people to understand.

I think this distinction matters well beyond Optivian. Early-stage companies often try to make one message do everything: positioning, category creation, SEO, paid acquisition, outbound and product explanation.

Those jobs do not always require exactly the same language.

Sometimes a narrow message is useful because it captures existing demand. Sometimes a broader message is necessary because it creates the right long-term perception. The difficult part is using the first without getting trapped inside it.

Sometimes the product has to explain the category

This leaves another problem.

If the product is broader than the familiar categories around it, how do you explain it without writing increasingly long paragraphs about what it is not?

The best answer we have found so far is surprisingly simple.

Show it.

Our end-to-end demo video has become one of the most effective ways of explaining Ollie.

I initially thought about the demo mainly as a product or conversion asset. Increasingly, I see it as a positioning asset.

Every word we use carries existing associations. “AI sales” carries one set of assumptions. “Sales assistant” carries another. “Deal intelligence” creates another mental model.

The more unfamiliar the product is, the more explaining it through category labels can sometimes make things worse.

A demo bypasses part of that problem.

Show Ollie understanding what has happened in a deal, identifying what is missing and helping the seller actually do the next piece of work, and the difference becomes much easier to grasp. People can understand the behaviour before they know what category to put it in.

That has been an important learning for me.

When the vocabulary for a new product is still developing, demonstrating behaviour can do more positioning work than adding another paragraph to the website.

It has also changed how I think about category creation.

“Creating a category” can sound like a branding exercise: come up with a name, publish a manifesto and put a new label on the website.

In practice, it feels much more operational.

You need to repeatedly help buyers connect three things:

the problem they already recognise → the limitation of the current way of solving it → the new way of thinking about it.

That happens across the whole GTM system. It affects which content you create, which keywords you target, which competitive comparisons you make, how sales explains the product, what the demo shows and even what you deliberately choose not to compare yourself with.

Visibility alone is not enough. If the market repeatedly sees you but files you under the wrong category, more reach can simply reinforce the wrong positioning.

At the same time, perfect positioning without distribution does not accomplish much either.

The difficult part is building awareness and understanding at the same time.

The category name may come later

I do not know whether AI deal execution becomes an established software category.

Perhaps AI sales co-worker becomes terminology buyers naturally understand. Perhaps the market eventually settles on completely different language.

I do not think we can fully control that.

What we can control is whether people gradually understand the distinction we are trying to make.

Not another AI tool for generating more activity.

Not another place to look at sales data.

Not another AI sales assistant waiting for the seller to prompt it.

The idea is AI that understands an active deal deeply enough to help move it forward.

For me, the more interesting marketing problem is how you get from something buyers already understand to something they do not yet have language for.

You have to borrow familiar categories without becoming trapped inside them. Capture demand that already exists while educating the market about a broader problem. Build visibility without reinforcing the wrong mental model.

And sometimes, when the words are doing too much work, stop explaining and show people what the product actually does.

That is probably the biggest lesson for me so far:

When you are building a category, positioning is not just deciding what to call yourself. It is managing the path from what the market already understands to what you want it to understand next.