Subscribe | Advertise

Good Morning,

Welcome to this special weekend edition of The Deep View, presented in partnership with Momentum.

By Jonathan Kvarfordt aka “Coach”
Head of GTM Growth Momentum
Founder of GTM AI Academy

Prompting Isn't the Problem.

Prompting Without Context Is. (for access to our 200+ prompt library, see below)

Over the past year, we’ve seen AI go from toy to tool, but most teams still use it like a guessing machine. They toss a half-written prompt into ChatGPT, paste in a call transcript, and hope the output sticks.

And then wonder why:

• CRM fields don’t match do not match the data

• Insights are vague

• Coaching feels generic

• Trust in the AI drops

This isn’t a tooling issue.
It’s a prompting + context issue.

And solving it starts with the design of your prompts.

Let’s Look at an Example

AI really thrives when it has something to mimic, or an example to follow. So the goal is to find experts in the field of whatever you do (and the expert maybe you!) that can give you examples of what good looks like.

For example, let’s use my friend Christian Krause (if you are in sales and not following him you should)

Now this is a screenshot from a Linkedin post he had around writing good cold emails. So then I took this and made it into a prompt.

Note the Hashtags markdown formatting and also that I literally took his email writing overview and made it into a prompt.

I gave the AI theory of what good looks like and then gave it an example to follow. Using both creates AMAZING results.

We humans learn best when we have 2 things: 1-Theory or context and 2-real live examples. And guess what this prompt has? You got it… Theory/context and live example to follow.

Quick tip, asking the AI to ask you questions to make the output better helps dramatically.

Then We Layer Context

Most people do not prompt with enough context, most of my prompts are VERY long because I am giving context and direction of how to use that context. 

However, prompts or context in the prompt alone aren’t enough. They only work if they run on relevant, structured context from other data sources.

If your AI doesn’t know:

• What the deal stage is

• What CRM fields exist, what is empty or context of the deal

• What methodology you're using (SPICED, MEDDPICC, etc.)

• What industry it is in or lead source

Then you are missing context and therefore your AI results will suffer and not be as accurate.

Most people forget to add this and therefore the AI is running on 1/8th of the context.

…then your perfect prompt still outputs noise.

Basic AI vs. Context-Engineered Prompting

Momentum is like a Data Refinery that cleans through multi stages of agentic-like workflows with validations happening across a company, so that CLEAN and trustworthy data is entered into the CRM. Momentum has a custom integration that can auto fill (or serve up info to your team to approve) 99% of all possible updateable field types in Salesforce. Most other companies can only do pick lists or text fields and they are leaving out a HUGE amount of data as a result.

The difference?
Context. Structure. Precision.

This is what we call context engineering and our data refinery and it’s the real unlock behind AI that actually works.

Momentum becomes a first party structured data provider with clean, unbiased, trustworthy data into your Salesforce CRM.

Want to Stop Guessing? Start with the Momentum Prompt Library

Momentum isn't just another AI tool. It's your AI-powered GTM execution layer.

To help you build this in your own org, we’re giving full access to the Momentum Prompt Library which are used by high-performing teams across Sales, RevOps, CS, Enablement, and Marketing.

🧰 Inside you’ll find:

• 200+ structured GTM prompts

• Context-enriched examples for SPICED, MEDDPICC, and more

• Prompt templates for deal reviews, call summaries, QBRs, CRM auto-fill, and Slack alerts

• Ready-to-run agents built into Slack + Salesforce

• Clean formatting, structured outputs, traceable insights

Access the library now

Final Thought

Prompting will always matter. But prompting with context is what powers Agentic workflows and will take your team and company to a whole other level.

The future of AI in GTM is structured systems that run on clean context, validated outputs, and trusted workflows.

That’s what we’ve built at Momentum.

And now, you can too.

Author Bio

Jonathan Kvarfordt is Head of GTM Growth at Momentum.io and founder of the GTM AI Academy. Recognized as a top enablement strategist by SEC three years running, he advises AI-first startups and leads GTM AI education at Bryant University. He believes prompts are systems, not syntax—and context is what turns AI into impact.