Market Size, Competition & Traction Slides
If Problem / Solution / Why Now decide whether the deck gets read, Market, Competition, and Traction decide whether it converts to a meeting. These are the slides where sophisticated investors and buyers stress-test whether the opportunity is real and whether you have the numbers to back it up.
AI is exceptionally good at these slides — if you use grounded research tools and resist the urge to inflate.
What You'll Learn
- The TAM / SAM / SOM bottom-up framework and how to prompt AI to build it
- The right way to do competitive landscapes (and the wrong way)
- The "10 things you did this quarter" prompt for traction slides
- Why grounded research tools (Perplexity, Gemini Deep Research) are non-negotiable here
The Market Slide
The most common pitch deck mistake is putting a giant industry number ("$1.2 trillion global software market") on the Market slide. This number is technically accurate and completely meaningless. Sophisticated investors discount it instantly.
The right structure is TAM / SAM / SOM, built bottom-up.
- TAM (Total Addressable Market) — the total revenue opportunity if every possible customer bought
- SAM (Serviceable Addressable Market) — the slice you could realistically sell to today
- SOM (Serviceable Obtainable Market) — what you can capture in 3-5 years given current resources
Bottom-up means you calculate from the number of customers times the realistic annual spend, not from "X percent of a trillion-dollar market."
The Bottom-Up Market Sizing Prompt
I am building a Market Size slide for a pitch deck. My ideal customer profile: [describe the buyer — size of company, industry, geography, role of buyer].
Build a bottom-up TAM/SAM/SOM model:
- TAM: estimate the total number of organizations globally that fit my ICP. Multiply by realistic average annual contract value (ACV). Show your math.
- SAM: narrow to the geographies and segments I can sell to today. Show your math.
- SOM: assume realistic market penetration in 3-5 years given seed-stage resources. Show your math.
For each number, cite the source you used (industry report, government statistic, public competitor numbers). Be explicit about every assumption.
The key phrase: "show your math." Without that, AI will pull a plausible-looking number and you will be unable to defend it when a VC asks where it came from.
Always Cross-Check With Perplexity
After ChatGPT or Claude gives you the bottom-up model, paste each number into Perplexity:
Verify this claim with web sources: "There are approximately [N] mid-market manufacturing companies in the United States with revenue between $100M and $1B." Return at least 3 sources and flag the highest-quality one.
Perplexity returns cited sources you can reference in the slide footnote. This is the difference between an investor saying "where did you get that?" and "I see your source — let's move on."
The Competition Slide
The classic 2x2 quadrant ("price vs features," with your logo in the top-right) is the most-mocked slide in venture. Every VC has seen 10,000 of them. They still work — but only if the axes are non-obvious and you are honest about competitors.
A stronger pattern in 2026 is the feature/job comparison table:
| Capability | Incumbent A | Incumbent B | Us |
|---|---|---|---|
| Onboards new vendor in under 5 days | No | Partial | Yes |
| Auto-fills SOC 2 questionnaires | No | No | Yes |
| Live for SMB and mid-market | SMB only | Mid-market only | Both |
| Built specifically for procurement teams | No (general workflow) | No (legal-led) | Yes |
This format tells the investor what the competitors cannot do — which is what you are betting your company on.
The Competitive Landscape Prompt
Build a competitive landscape for my company. My product: [1-2 sentences]. The main jobs to be done that I solve: [list 3-5 jobs].
Steps:
- List the 5-7 most credible competitors in this space today (named companies, not categories). For each: 1-line description, last known funding stage, rough employee count.
- For each competitor, score how well they do each job to-be-done: Yes / Partial / No.
- Identify the 2-3 jobs where I am clearly better — these will become our differentiation axes.
- Identify the 1-2 jobs where I am clearly worse — what is the honest answer to "why don't customers just use X instead?"
The "honest answer to why don't customers use X instead" is the prompt addition that distinguishes mature founders from hopeful ones. Investors ask this question; have the answer ready on the slide.
Verify Competitors Are Real
AI will sometimes hallucinate companies that do not exist. Always cross-check with Perplexity or a direct web search:
Confirm the following companies exist and are active as of 2026, with their last known funding stage. If any do not exist or have shut down, flag them: [paste list].
The Traction Slide
Traction is the only thing on a pitch deck that cannot be faked. It either exists or it does not. AI helps you present traction well — it does not generate the traction itself.
The classic mistake is putting one big revenue number on the slide. The right pattern is to show momentum: a chart where the right-hand side is much higher than the left.
The Traction Slide Prompt
Help me decide what goes on my Traction slide. Below are 10 things I did this quarter:
[paste 10 bullets — customers signed, revenue closed, retention numbers, pilots, hires, product launches, press, partnerships, etc.]
Pick the 3-5 strongest traction signals. For each: (1) why this signal matters to a Series A VC, (2) the exact metric I should display, (3) the chart type that makes the momentum visible, (4) the action title that frames the chart.
Also flag any of my 10 items that I should NOT put on the slide — and explain why those would actively hurt.
The "flag what to leave off" part is critical. Founders constantly put weak signals on traction slides (a single LOI, a vanity press mention, a hire). AI helps you cut them.
Common Traction Charts That Work
- Monthly recurring revenue: bar chart, last 6-12 months, ideally going up and to the right
- Logo wall: 6-12 customer logos with at least one large name
- Cohort retention: cohort analysis showing newer cohorts retain better
- Pipeline: total qualified pipeline + close rate trend
- Engagement: DAU/MAU ratio, or sessions per user over time
For each, ask AI:
Suggest the action title for a [MRR / retention / logo wall] chart that shows [describe data]. The title should be a complete sentence stating what the chart proves. Maximum 14 words.
When You Do Not Have Traction Yet
A common situation: pre-product or pre-revenue startups. The right play is to swap "Traction" for "Validation":
I am pre-revenue. Help me build a Validation slide instead of a Traction slide. Inputs:
- Customer discovery interviews: [N] conducted
- LOIs / design partners signed: [list]
- Pre-orders or paid pilots: [list]
- Founder-relevant data points: [e.g., I sold the previous company to X for $Y]
Structure the slide as 3-4 quantified validation signals with one supporting visual.
Investors know what pre-revenue looks like. Honest validation beats inflated traction every time.
Key Takeaways
- TAM/SAM/SOM must be built bottom-up — number of customers times ACV, with the math visible
- Always cross-check market numbers in Perplexity or another grounded research tool
- Competitive slides should use a job/feature comparison table, not the classic 2x2
- Include the "why don't customers just use X" question in every competitive prompt
- Traction slides need momentum, not one big number — and you should also ask AI what to leave off
- Pre-revenue companies use a Validation slide with LOIs, pilots, and discovery counts