Claude Docs and Slides for Sales Teams: The Deck Tax Is Dying
In mid-September 2026, Anthropic quietly killed Claude Cowork as a separate mode, folded it into the main Claude app, and shipped two new surfaces in beta: Claude Docs and Claude Slides. Most of the coverage framed this as an office-suite play against Google and Microsoft.
That's the wrong frame for anyone who carries a quota. The right frame: the most expensive non-selling activity in sales — building documents and decks — just moved inside the AI conversation. Claude's output is no longer text you paste somewhere. The deliverable itself is the output.
We've been running Claude in production GTM workflows since early 2025 (start with the complete guide to Claude for SDRs if you're new). Here's what actually shipped, why it matters more for sales teams than for the "AI office suite" crowd, and five workflows worth stealing this week.
What shipped, in 30 seconds
- One Claude. You no longer pick "Chat" or "Cowork" up front. Claude decides whether your request needs a quick answer or longer-running agentic work. Existing Cowork chats, projects, artifacts, connectors, and skills carry over.
- Claude Docs (beta, paid plans). Create and edit documents inside a conversation. Real-time co-editing via shareable links, on desktop and mobile. Exports to Microsoft Word or Google Docs.
- Claude Slides (beta, paid plans). Editable presentations you can present directly from Claude, or export as PowerPoint or PDF.
- Rollout: Pro and Max first, staged over the coming weeks across web, desktop, and mobile. Team and Free plans follow. Enterprise orgs get at least 30 days' notice before changes.
- Claude Code stays separate — it remains the terminal/IDE product for engineering work.
- Agentic actions still ask permission by default, with autonomous modes that check in when review is needed.
If you're weighing Claude against the alternative for this kind of work, our Claude vs ChatGPT for sales teams breakdown covers the underlying model differences; this post is about the new surfaces.
The deck tax, quantified
Every AE knows the routine: discovery call ends, and now you owe the prospect a recap doc, a tailored deck for the champion to forward internally, and maybe a mutual action plan. None of that is selling. All of it eats selling hours.
The pre-Docs workflow with AI already helped, but it had a copy-paste seam running through it:
| Step | Before Docs/Slides | After |
|---|---|---|
| Draft proposal | Prompt Claude, paste into Google Docs, fix formatting | Claude produces the doc; you edit in place |
| Build the deck | Prompt for an outline, rebuild it slide by slide in PowerPoint | Claude builds the deck; export PPTX or present from Claude |
| Champion feedback | Email attachments back and forth | Shareable link, real-time co-editing |
| Version control | "Proposal_v4_FINAL_v2.pptx" | One living document in the conversation |
The seam mattered more than it looked. Reformatting is where the time went, and it's why most reps still hand-built decks even with a Claude Pro subscription. That seam is now gone — with caveats we'll get to.
Five workflows to steal
1. Discovery call → recap doc before the prospect's next meeting
Pipe your call notes (or transcript) into Claude and ask for a one-page recap doc: their words for the problem, quantified impact, proposed next steps. Export to Google Docs or share the link directly. Speed is the differentiator here — a recap that lands 30 minutes after the call reads very differently than one that lands Thursday. The prep side of this is covered in our AI sales meeting prep guide.
2. The forwardable deck
Your champion doesn't buy — they sell internally on your behalf. Build them a 5-7 slide deck in their company's language: their initiative names, their metrics, their org's priorities. Prompt pattern that works:
"Build a 6-slide deck a Director of Sales Ops would forward to their CRO. Their initiative is called [X]. Frame our solution as accelerating [X], not replacing anything. One idea per slide, no jargon."
Then export PPTX and let the champion re-skin it in corporate template if they need to.
3. QBR decks from account data
QBR season used to mean AEs disappearing for a week of deck-building. Feed Claude the account's usage data, support history, and goals from the last QBR, and have it produce the first full draft — trend narrative, wins, risks, expansion asks. With long-context workflows you can hand it a quarter's worth of raw material in one shot.
4. Mutual action plans as living documents
MAPs die in attachments. As a shared Claude Doc with a link, the plan becomes the single artifact both sides edit — dates, owners, dependencies. When the buyer updates a milestone, you see it. When they stop touching it, that's signal too.
5. Persona one-pagers at scale
SDR teams need leave-behinds per persona, per industry. This was never hard to write — it was hard to format twenty times. Now it's a loop: one master prompt, one doc per persona, consistent structure. Pair it with automated lead research so each one-pager cites the pains that segment actually reports.
The caveats (beta means beta)
- Paid plans only, staged rollout. If you don't see Docs/Slides yet, that's the staging, not you. Team plan admins are waiting on the follow-on rollout.
- Brand control is the open question. Exported decks will need your template applied for anything customer-facing at enterprise. Treat Slides as the content pass; brand polish still happens in PowerPoint for now.
- Data hygiene still applies. A shareable co-editing link is convenient right up until someone pastes another customer's numbers into a QBR deck. Same rules as any AI tool: no cross-customer data in one artifact, and check your org's retention settings before sharing links externally.
- This doesn't replace judgment. Claude will happily produce a confident deck built on a wrong read of the account. Garbage in, beautifully formatted garbage out.
Where this fits in your stack (and what it doesn't solve)
Docs and Slides compress the asset step of selling. They do nothing for the step before it: knowing who to build the asset for and why now.
A perfect forwardable deck aimed at an account that isn't in-market is still zero revenue. The order of operations that works: signal first (who's on your site, who changed jobs, who's researching the category), play second (what action that signal demands), asset third (the doc or deck that executes the play). That first two-thirds is what MarketBetter does — it tells your SDRs who to target and exactly what to do, so tools like Claude Docs execute plays instead of producing beautifully formatted guesses. If you're building that muscle, start with our SDR playbook template guide.
And if your team is further along the automation curve, the same "asset as output" logic extends to LinkedIn workflows and even forecasting — the pattern across 2026 is consistent: AI stops giving you text about the work and starts giving you the work.
Bottom line
- If your reps have Claude Pro/Max: turn on Docs/Slides the moment it appears, and standardize the five workflows above. The recap-doc and forwardable-deck plays alone are worth hours per rep per week.
- If you're on Team: it's coming; draft your prompt library now.
- If you're evaluating AI for GTM broadly: this release is another data point that the chat-window era is ending. Assets, actions, and agents are the output now — a shift we've tracked since Claude first connected to real sales tools.
The deck tax funded a lot of late nights. Few will mourn it.
Want the "who and why now" half of the equation solved too? MarketBetter identifies your website visitors, tracks buying signals, and hands your SDRs a daily playbook of exactly what to do. Book a demo →