Q3 2026 Indie Publisher Diligence Shifts to Demo Evidence Packets - Industry Analysis for Micro-Studios

Pitch decks did not disappear in 2026. They stopped closing rooms alone.

Across Q3 2026 partner intake conversations—late-August windows for many late-2026 and early-2027 launches—micro-studios report a repeated ask: “Send the demo evidence packet.” Not a wishlist screenshot. Not a vertical slice label. A mapped folder tying build ID, store copy, playable scope, and operating cadence to artifacts a partner can open in five minutes.

This Industry analysis reads that shift through practical lenses for 1–10 person teams: what changed, why May 2026 is preparation month, and how it connects to release-evidence taxonomy, milestone checklists, and the fest stack you already built for players.

Why May 2026 is the preparation window

Partner intake for late-2026 launches clusters August–September 2026 in shared forum timelines—not because of a single public calendar, but because:

• Fest demos lock July–August
• Due diligence runs parallel to fest marketing, not after
• Teams that scramble post-fest submit packets with stale build IDs

Micro-studios treating May as “still pre-production” miss that diligence is concurrent with fest prep, not sequential after wishlists spike.

What changed (2025 → Q3 2026)

Three industry-level pressures converged:

1. Post-hype funding discipline — After 2023–24 reset, partners weight executable scope over narrative ambition. Decks still sell vision; packets sell credibility.
2. Dual-SKU and demo density — PC fest builds plus browser slices (Opinion on unscoped web) multiply surfaces where lies surface. Diligence follows player complaints.
3. Cert and platform spillover — Console reproducible-capture expectations (Quest resubmission playbook) trained partners to expect bullet → file maps, not Slack threads.

The shift is process, not a single publisher memo. Forums, accelerator office hours, and post-mortems align: teams that show up with folders close faster.

Signals observers cite (qualitative, 2026)

Observers across public dev forums and conference hallway threads repeatedly mention:

• Partners asking “where is the evidence folder?” after trailers
• Accelerators rejecting cohort apps with strong art but no operating cadence
• Due diligence calls starting with README review, not pitch narration
• Follow-up meetings stalling when zip paths break

This article does not quantify close rates—it describes a directional industry habit micro-studios should plan for.

Direct answer (TL;DR)

Prepare in May–July 2026 so August intake is submission, not scavenger hunt.

What you will have after May prep

A versioned zip a stranger can navigate in five minutes, tied to a real build hash and store claims—a receipt that your studio executes like cert-bound teams, even if you never sign a publisher deal.

Who this analysis is for

• Founders entering first publisher or accelerator diligence
• Teams with October 2026 Next Fest demos who may pitch partners the same quarter
• Studios that rationalized stacks and need one credible evidence root

Not for: established publishers’ internal process documentation—we analyze micro-studio reception, not partner HR.

Anatomy of a demo evidence packet (2026 consensus pattern)

Synthesized from shared 2025–26 forum patterns and accelerator checklists—not one named publisher policy:

demo-evidence-packet-2026-05/
  README.md                 # one-page map
  01-build/                 # hash, export preset, engine version
  02-store-truth/           # truth audit inventory + scope card
  03-visual-stack/          # capsule/screenshot gate PASS table
  04-playable/              # demo keys, itch scope, heap log if web
  05-operations/            # four weekly review sheets
  06-milestones/            # payment checklist status (if deal active)

Partners are not asking for novelty—they are asking for findability.

README map (highest-signal page)

# Demo evidence packet - <Game> - 2026-05-16

## Build under review
- ID: 1.0.4.512
- Platform: PC Steam demo
- Fest target: October 2026 Next Fest

## Claims vs proof
| Claim on store page | Proof path |
|---------------------|------------|
| "45 min demo, floors 1-6" | 02-store-truth/scope-card.md |
| "Co-op not in demo" | 02-store-truth/truth-inventory.md row 7 |
| "Prototype not vertical slice" | 02-store-truth/demo-label-screenshot.png |

## Operating cadence
- Last four Friday sheets: 05-operations/

## Contact
- Owner: <name>

Five-minute read test: if a stranger cannot navigate, the packet fails before content is judged.

How this ties to release-evidence/

Your repo-level release-evidence/ tree is the factory; the diligence packet is the export for partners.

Do not duplicate legal contracts—reference offline with row in README.

Operating review as diligence artifact

The 30-minute operating review stopped being “internal productivity” and became receipt evidence that a team can steer without heroics. Four consecutive Friday sheets in 05-operations/ answer:

• Do they know their burn of scope?
• Do marketing and engineering see the same numbers?
• Is Block 4 honest about store truth?

Partners treat cadence as proxy for ship probability.

Milestone deals under the same lens

If a term sheet exists, milestone payment checklists belong inside the packet as 06-milestones/—same bullet-map discipline as cert resubmissions. Mixed messages (“we are milestone-driven” + empty acceptance rows) kill trust faster than missing art.

Section-by-section build notes (expanded)

01-build/

Include:

• ProjectSettings version string screenshot
• Export preset name for fest demo
• Git tag or commit hash
• Known issues list (honest, short)

Partners reproduce from hash—without it, walkthroughs are theater.

02-store-truth/

Merge outputs from:

• 7-day truth audit inventory
• Demo label screenshots
• itch scope card if web SKU exists (browser opinion)

One row per claim; zero “TBD” rows at send time.

03-visual-stack/

Export:

• Composition gate scorecard
• Phone-width grid PNG from 14 tools workflow
• Capsule iteration date from calendar article

Visual diligence is “does marketing match build,” not “is art pretty.”

04-playable/

• Steam demo keys or instructions
• itch link only if green checklist complete
• Save file policy (disabled / enabled) stated plainly

05-operations/

Four Friday operating review sheets minimum—Block 4 marketing truth visible.

06-milestones/ (conditional)

If negotiating: paste checklist state from milestone article—accepted vs pending rows, not promises.

Live walkthrough vs packet (how calls changed)

2024 call: share screen, improvise, promise roadmap dates.
2026 call: partner read README beforehand; screen share confirms rows, does not discover them.

Prepare talk track as row confirmations:

“Row 4 scope card matches build 512—opening 02-store-truth/scope-card.md.”

Improvised feature demos without inventory rows undermine the packet you sent.

Fest marketing stack enters diligence

Partners watch player-facing surfaces:

• Truth audit week outputs
• Screenshot composition gates
• Godot WASM trend if browser SKU exists

A crashy browser link in the wild undermines a pristine PC packet—industry feedback loops are public.

Regional and platform nuance (EU vs US framing)

EU-facing partners increasingly ask privacy and DMA-adjacent distribution questions alongside demo proof—cross-read DMA practical analysis as annex, not core packet.

US-facing partners weight visit-to-wishlist conversion and fest timing—cross-read capsule iteration calendar.

Packet stays one folder; README links annexes.

Publisher vs accelerator vs platform holder (how asks differ)

One packet skeleton serves all—tune README emphasis, do not maintain three zips.

Red flags partners report (synthesized patterns)

None require invented studio names—patterns repeat in open office hours.

Green flags (inverse patterns)

• README completes in under five minutes with intern test
• Truth inventory matches vertical slice honesty labels
• Screenshot PASS table included from composition gates
• Web SKU has heap log or explicit “PC-only” scope card
• AI disclosure touchpoints listed per checklist

Ninety-minute packet bootstrap (May sprint)

If co-founder cannot find “co-op claim” row in ninety seconds, revise README.

Zip hygiene and security

• No .env, keys, or player PII in repro zips
• Redact Discord names in screenshot annotations
• Version zip filename: game-demo-evidence-2026-05-16.zip
• Password optional for embargoed builds—password in separate channel

Size target: under 100 MB for email-friendly sends; link to cloud for videos.

Relationship to AI-assisted workflows

Partners increasingly see generative assist in art and copy. Packets should include ai-disclosure/ summary per platform checklist—not because every partner mandates it, but because omission reads as hiding. Assistive vs generative taxonomy belongs in 02-store-truth annex.

Wishlist metrics without fantasy forecasts

Include Steamworks export with dates and methodology note:

“Visit→wishlist rate measured Fridays; 48h lag; no forecast implied.”

Pair with top-18 conversion tools discipline—charts without methodology read as deck theater.

Console and PC parallel tracks

Teams running Quest resubmission discipline should add optional annex 07-console/ with bullet map—publishers evaluating PC-first SKUs still like seeing console risk managed.

Stack rationalization as diligence narrative

Founders who collapsed engines and storefronts should state it in README:

“Primary SKU: PC Steam. Secondary: scoped web demo per opinion/policy. Deprecated: mobile port exploration Q1 2026.”

Rationalization is a risk reduction story when evidenced, not just a blog post.

Contrarian micro-studio responses (industry reactions)

Some teams push back:

• “Packets are big-studio process.”
• “We are art-led, not ops-led.”
• “Our game speaks for itself.”

Counter-pattern from forums: art-led teams with minimal README maps still close faster than ops-heavy teams with chaotic Google Drives. Minimal structured beats maximal chaotic.

EU AI Act and privacy annex (light touch)

Not legal advice—industry observation: EU-facing funds ask privacy template version dates alongside demo proof. Link 14 privacy templates as annex row—do not paste full policies into zip unless counsel approves.

Post-intake maintenance

After August send:

• Update packet tag when build ID changes
• Append operating sheets weekly
• Re-run truth inventory if store strings change

Stale packet worse than no packet on follow-up call.

Solo founder vs three-person team packets

Solo founders skip 05-operations at peril—substitute biweekly sheets if true weekly impossible, note cadence honestly.

Platform feature applications (non-publisher)

Steam featuring, itch front-page requests, and curator outreach increasingly resemble lightweight diligence—same README skeleton, shorter zip.

Contrasts with 2024 “deck-only” accelerators

Accelerators that once accepted slide PDFs now request playable + folder before cohort start. The industry analysis trend is horizontal across publisher types, not publisher-exclusive.

Risk disclosure without fear-mongering

Packets should include known limitations section:

“Web SKU: floors 1–6 only. Co-op: PC full game, not in demo. Performance: target 60fps 1080p min spec.”

Honest limitations increase close rates in synthesized forum patterns—partners model risk, not hope.

Read order for sibling articles (May week)

1. This industry framing
2. Release-evidence taxonomy
3. Truth + visual + web siblings as needed
4. Operating review cadence
5. Export packet zip

Skip step 5 and diligence stays theoretical.

What diligence is not asking for (May 2026)

• Invented revenue multiples from unnamed “similar titles”
• Fake team alumni pedigrees
• Guaranteed wishlist forecasts
• Legal advice in README

Stay in evidence lane per site policy—pattern playbooks, not named-studio biographies with invented metrics.

Timeline table (May–October 2026)

Decision tree

Publisher conversation scheduled?
  no → Still build packet skeleton
  yes → Cold-open test passed?
        no → Fix README paths
        yes → Send zip + offer live walkthrough

Two-team diligence drill (recommended)

Before any partner call:

1. Partner role: co-founder opens zip cold for five minutes, lists three questions.
2. Founder role: answer only using README paths—no live demo yet.
3. Swap: repeat with build running.

Gaps found in step 1 are cheap; gaps found on the call are expensive.

Funding climate context (without invented stats)

Industry commentary in 2025–26 consistently describes smaller checks and more verification steps per check—not a single dataset, but a directional climate where packets reduce perceived risk per dollar asked. Micro-studios cannot change macro funding; they can change perceived execution risk with folders.

Integration with price and monetization talks

If diligence coincides with price-anchor worksheet work, include pricing-and-store/ annex with locked anchor doc version—partners connect monetization discipline to ship discipline.

Key takeaways

1. Q3 2026 diligence centers demo evidence packets, not decks alone.
2. May–July is preparation; August is intake for many partners.
3. README bullet maps are the highest-signal artifact.
4. release-evidence/ feeds packet exports.
5. Operating review sheets prove cadence.
6. Store truth and visual stack are diligence surfaces.
7. Browser SKUs need heap logs or honest absence.
8. Milestone checklists integrate if deals are live.
9. Public fest marketing can invalidate private packets.
10. Micro-studios with folders close faster—forum pattern, not guaranteed outcome.
11. Stakeholder type tunes README emphasis, not folder chaos.
12. Red flags are structural—broken paths, stale sheets.
13. Ninety-minute bootstrap beats August all-nighter.
14. Zip hygiene and size matter for email sends.
15. AI disclosure annex increasingly expected.

FAQ

Is this every publisher?
No—pattern synthesis from 2025–26 shared reports.

Do we send the whole repo?
No—curated zip with README map.

Accelerator vs publisher difference?
Accelerators weight team cadence more; publishers weight SKU truth more—same packet skeleton.

We have no publisher yet.
Build packet anyway before fest; reduces August scramble.

How is this different from milestone article?
Milestone piece is contract execution; this is pre-deal diligence framing.

Can we use Notion instead of zip?
Export PDF snapshots weekly; partners archive email zips.

Do influencers count as diligence?
No—unless contractually a platform partner; separate folder.

We are self-publishing only.
Packet still helps platform feature applications and fest curators.

Will this guarantee funding?
No—industry analysis describes patterns, not outcomes.

How often to refresh the zip?
After every demo build ID change or store copy freeze break.

Can lawyers review the packet?
Yes—send README + scope sections before full zip if embargo concerns exist.

Closing industry read (May 2026)

The indie market did not become hostile—it became verifiable. Decks still inspire. Packets prove you can ship what the deck inspires. Micro-studios that internalize Q3 2026 diligence as demo evidence culture—not a one-off PDF for a single call—enter fest season with calmer founders, fewer public trust fractures, and faster partner threads when funding windows open.

That culture starts in May with a README and a build hash—not in August with a prayer and a Drive link.

Packet version table (maintain in README footer)

Partners compare versions on follow-up calls—version discipline prevents “which zip was that?” confusion.

Email send template

Subject: <Game> — demo evidence packet v1 (build 1.0.4.512)

Body:

Attached is our demo evidence packet for the August conversation. Start with README.md (five-minute map). Build 1.0.4.512 matches our Steam fest demo branch. Happy to walk rows live on the call.

Short, pointer-first, respectful of partner time.

Common packet mistakes (May 2026)

1. Sending deck PDF instead of packet zip for “convenience”
2. README claims features removed in yesterday’s build
3. Hiding known demo crashes instead of listing them in 01-build/issues.md
4. Mixing milestone legal PDFs into playable folder without counsel review
5. Forgetting to update version table after patch
6. Using marketing superlatives in README instead of verifiable rows

When packets replace pitch competitions

Some 2026 juried events request evidence of execution before art awards—same skeleton applies. Jurors are not publishers, but they imitate publisher habits when evaluating ship risk. Treat festival jury packets like publisher zips: README first, binaries last.

Final one-liner for team standups

“If a partner opened our zip cold today, would they trust build 512?”

If the answer is no, May work is not done—regardless of wishlist count.

Close: Q3 2026 rewards studios that treat diligence like cert intake—mapped, boring, and true. Build the packet in May so your August is a send button, not a panic search through Discord files. The folder is the product; the meeting is the confirmation. Ship the zip before you ship the slide deck. Re-run the cold-open test after every milestone patch and after every store-copy change.