Automating the 'Un-Automatable': High-Fidelity Sales Decks via Gamma and Figma
TL;DR — Custom sales decks have always been the bottleneck nobody wanted to admit. With Gamma handling rapid AI-generated structure and Figma handling on-brand polish, triggered automatically from a CRM deal stage change in Make.com, the wait goes from two weeks to two hours. Sales teams actually use them. That changes the pipeline math.
The Two-Week Deck Tax
Sales teams used to wait two weeks for a custom deck. Sometimes longer.
The request came through Slack or a shared tracker, landed in the PMM queue behind three other priorities, got worked on when someone had a window, went through two rounds of revision to get the branding right, and arrived — finally — either just in time or just too late to matter in the deal.
This was not anyone’s fault. It was a capacity problem dressed up as a process problem. Custom sales decks require design effort, strategic judgment, and an understanding of the specific deal context. All three of those inputs are finite. When sales volume exceeds PMM capacity to produce custom assets, the queue grows and the wait grows with it.
The answer most teams reach for is templatisation: give sales reps a master deck and a PowerPoint they can customise themselves. Every PMM knows how this ends. The “customised” deck is either the master deck with the company name changed, or a frankenstein assembly of slides that looks like it was made by someone who last opened PowerPoint in 2018. Neither version closes deals.
There is a better answer, and it requires rethinking what automation can reach.
Why Design Has Always Been the Automation Ceiling
Content automation is mainstream at this point. Blog posts, email sequences, social copy, battle cards — the tools for generating these at scale are well-established and most PMM teams have experimented with at least some of them.
Design has stayed stubbornly manual. The reason is that design quality has historically required a human with visual judgment, brand knowledge, and the ability to make the hundreds of small decisions — layout, hierarchy, white space, colour application — that make a slide deck look considered rather than generated.
Gamma changes the structural layer of that equation. Figma changes the polish layer. Together, with a Make.com trigger connecting the two to your CRM, they change what is possible.
How the Gamma + Figma + Make.com Flow Works
The Trigger: CRM Deal Stage Change
The Make.com scenario watches your CRM for a specific deal stage change — typically “Proposal” or “Solution Review,” the moments when a sales rep needs a tailored deck most urgently. When a deal hits that stage, the scenario fires.
It pulls the relevant deal data: company name, industry, deal size, primary use case, the rep’s name, the key contacts on the buying team, and any notes attached to the deal record. This is the raw context the deck generation will work from.
The Structure Layer: Gamma
Gamma’s API accepts a structured prompt and generates a complete presentation with slides, layout choices, and content populated from your input. This is where the blank-page problem is solved in seconds rather than hours.
The Make.com scenario constructs a Gamma prompt from the CRM deal data: “Generate a six-slide sales presentation for [Company Name], a [industry] company with [size] employees. The primary use case is [use case]. The key challenge they are solving is [challenge from deal notes]. Include: a business challenge slide, a solution overview, a product capability summary, a relevant case study, a commercial summary, and a clear next step. Tone should be direct and peer-level, not vendor-to-buyer.”
Gamma returns a structured deck — good content, coherent flow, readable layout. It will not be on-brand. That is intentional. Gamma handles structure. Figma handles brand.
The Brand Layer: Figma
The Gamma output is mapped to a Figma template via the Figma API. Your brand template — the one your design team built, with the right typography, the right colour palette, the right slide layouts — is already in Figma. The automation populates that template with the Gamma-generated content.
This step requires upfront investment: working with your design team to build a Figma template that is structured for programmatic population, with clearly labelled content blocks that the automation can target. That investment is made once and serves every deck generated thereafter.
The output is a Figma file that looks like your brand because it is your brand template, populated with content that is specific to this deal because it was generated from this deal’s CRM data.
The Delivery Step
Make.com exports the Figma file as a PDF or shareable link and posts it to the sales rep’s Slack DM along with a brief summary of what was generated. The rep gets notified within two hours of the deal hitting the trigger stage — often faster. They receive a deck that is on-brand, deal-specific, and ready for a quick review pass before the next call.
Review time for the rep: five to ten minutes to check the content and personalise the opening slide. Total time from deal stage change to polished first draft: under two hours.
Where Human Judgment Still Lives in This Flow
The automation handles structure and initial content. It does not replace the rep’s understanding of the relationship, the nuance of the buying dynamic, or the strategic judgment about which angle to lead with in a specific deal.
What it eliminates is the two-week wait, the production bottleneck, and the blank-page problem for the rep who needs to walk into a meeting with something credible. A five-minute review and personalisation pass by someone who knows the deal is infinitely better than a two-week PMM production cycle — and it gives the PMM function leverage to focus on the strategic work rather than the production work.
This is Product-Led Growth logic applied to internal sales enablement: reduce friction, accelerate time to value, and let the rep spend their time on the interaction rather than the asset.
The Upfront Work Worth Doing
Two investments make this flow possible and both are worth being honest about:
First: the Figma template. This is a two-to-three-day design project with your designer to build a presentation template that is structured for programmatic content population. It requires design and PMM working together on slide architecture. It is the kind of work that feels like overhead until the first time the automation uses it, at which point the investment pays back immediately.
Second: the CRM data quality. The automation is only as good as the deal data it works from. Deal notes that say “follow up” produce unusable deck content. Deal notes that describe the challenge, the use case, and the key contacts produce decks worth reading. This is an alignment conversation with your sales team, framed not as “fill in more fields” but as “the better your deal notes, the better the first-draft deck you get.”
Key Takeaway for PMM Leads
High-fidelity design has been the ceiling of content automation because it requires both structural logic and brand judgment simultaneously. Gamma and Figma separate those two concerns — structure first, brand second — and Make.com connects them to your CRM so the trigger is always the right deal moment, not a manual request. The result is a custom sales deck delivered in hours, not weeks. The upfront investment in a Figma template and deal data hygiene is real. The return on that investment, measured in deals where the rep showed up with the right asset at the right time, is realer.
Frequently Asked Questions
Does Gamma produce presentation-quality content or just a rough outline?
Gamma produces well-structured slide content with coherent narrative flow — closer to a solid first draft than a rough outline. The quality of the content depends significantly on the quality of the prompt it receives, which is why the CRM data enrichment step is important. With good deal context, Gamma’s output is typically good enough to be useful after a light review pass. The brand layer is handled by Figma, so Gamma’s visual output is intentionally treated as placeholder rather than final.
What if our team uses Google Slides or PowerPoint rather than Figma?
The Figma approach works best because the Figma API supports programmatic template population more cleanly than Google Slides or PowerPoint. That said, a simplified version of this workflow using Google Slides via the Google Slides API is achievable if your team is already operating in Google Workspace. The structural generation step with Gamma remains the same; the brand layer becomes a Google Slides template population rather than Figma. You lose some design fidelity but gain integration simplicity.
How do we handle deals that need significant deck customisation beyond what the automation produces?
The automation is designed to produce a strong first draft, not a final deck for every situation. For deals with unusual requirements — a particularly complex competitive situation, an enterprise buyer with specific procurement constraints, a proof-of-concept scope — the automated first draft still saves significant time even if meaningful manual work follows. The PMM is starting from a structured, on-brand asset rather than from nothing, which changes the production time even when the customisation work is substantial.
Can this workflow handle different deck templates for different market segments or personas?
Yes, and this is one of the more powerful extensions of the base flow. Make.com can route the Gamma prompt and Figma template selection based on deal attributes — company size, industry, or primary use case — so that an enterprise deal triggers a different template and prompt structure than a mid-market deal. Building this routing logic requires a bit more scenario complexity upfront, but it means the automation is not just fast — it is segment-aware from the first draft.