New Ebook! AI 101 for MedTech Sales: A Practical Guide (with Prompts You Can Copy)

Commercial leaders: Your 2027 budget should include a line item for AI usage. Here’s why.

Using AI is no longer optional for MedTech reps; it’s essential to staying competitive. Commercial leaders must budget accordingly.

November 16, 2026

When we teamed up with MedTech Dive to survey 150 medical device sales reps on their AI usage, we found that AI users are 3x more likely to hit quota.

It’s no wonder why: By using AI, reps can significantly reduce the amount of time they spend researching prospects and creating call plans, allowing them to pursue a higher volume of opportunities (and walk into each conversation with a high-quality talk track).

For commercial leaders, this means using AI is no longer optional. If your reps don’t use AI, they’ll lose ground in their territories to competitors who do.

With that said, there’s something you must keep in mind as you embark on your AI transformation: For you as a customer, buying AI is fundamentally different from buying traditional software.

Let’s explore why that’s the case, and what it means for you as you evaluate AI vendors.

The usage-based model of AI pricing, explained

Many AI vendors employ usage-based pricing, meaning they charge you for every instance of usage. The reason they do this is inference costs. Every time you tell an AI tool to complete a task for you, the vendor incurs inference costs; the more demanding the task, the greater the inference costs. Usage-based pricing is what makes it possible for AI vendors to stay in business.

Contrast this with traditional software, where seat-based pricing is the norm. Traditional software vendors don’t have to worry about inference costs. When you add information to a lead record in your CRM, for example, the CRM vendor doesn’t incur any additional cost. To a CRM vendor, the cost of serving a user who updates 20 records weekly is about the same as the cost of serving a user who updates 200 records weekly, which is why the vendor has no problem charging the same flat fee for each user’s seat regardless of how much they use the product.

From your perspective as the customer, the key difference between seat-based pricing and usage-based pricing is predictability. If your CRM vendor charges $100 per seat, you can predict what it will cost to add a new rep to your account: $100. Meanwhile, predicting the cost of giving that same rep access to an AI tool is more difficult. Why? Because the cost depends on the rep’s usage of the tool. The cost could be $100 per year, or it could be $1,000 per year.

AI vendors know that this lack of price predictability is a barrier to doing business, and in many cases will try to eliminate the barrier by selling a set amount of usage for an upfront fee. You might, for example, pay a $10 upfront fee in exchange for 100 credits of AI usage.

The benefit of this approach is that it eliminates risk for you as the customer. You don’t need to worry about being surprised by a large bill. You paid $10 up front, and now you’re free to use the tool until you reach your 100 credit limit.

Which leads to an important question: What is a credit?

The basics of AI credits

One way to think of a credit is as a unit of work. More demanding tasks require more units of work, and therefore consume more credits. Having AI draft an email may cost you 1 credit, whereas having it create a 20-slide deck may cost you 5 credits. (Please note that these credit amounts are hypothetical and purely for illustrative purposes.)

This pre-paid credit approach has become quite common amongst AI vendors as a way to give customers the price predictability of traditional software. That said, it’s not always perfect predictability, as two AI workflows that seem similar can consume different amounts of your allotted credits. The reason? It comes back to inference costs, the costs that AI vendors incur when generating outputs like emails and slide decks.

To understand why, imagine two users of the same AI tool: Rep A and Rep B. Both want the tool to draft an email for them. Rep A wants the tool to draft a brief, generic cold email to a physician; Rep B wants it to draft a detailed response to a prospect’s list of technical product questions.

In both cases, the output is an email, but the inference costs incurred by the vendor will be different (due to the difference in the amount of work the tool needs to do). This is why the vendor is unable to charge a universal flat fee of, say, 1 credit per email; there’s too much risk that the revenue from credit sales will fall behind inference costs, ultimately creating financial problems for the vendor.

In short: The variability of inference costs is the reason AI vendors can’t always definitively tell you what you’ll be able to achieve with your allotted credits.

What this means for MedTech commercial leaders evaluating AI vendors

Again, using AI is no longer optional; it’s essential to staying competitive in your market. As we’ve explained here, buying AI is fundamentally different from buying traditional software, and we recommend having a dedicated AI usage line item flexible enough to expand as the technology advances and your reps use it more and more.

The size of the line item should reflect your expectations for next year. Will you be expanding into new markets? Commercializing new devices? Employing new sales motions or methodologies? Taking on additional funding? These kinds of changes have implications not only for team size but also individual AI usage, and therefore should be considered, such that you feel confident you’re accounting for the right amount of overall AI usage.

We also recommend working exclusively with consultative AI vendors, i.e., AI vendors that give you proactive, hands-on account managers who will make sure your reps’ AI usage is productive and cost-effective, and even help with budgeting (see previous paragraph). This is the best way to maximize the return on your AI investment.