CrewChat
CrewChat
Solving Impact's retention gap by bringing on-set communication into a mobile app that 200+ crew use on every production.
+380% time spent in the app · 4 studio deals · 110+ productions onboarded
The One Workflow Worth Solving
Most coordinators and crew already used Impact to network and get hired. Once they landed on a project, which could run for months, they had no reason to open Impact again.
We needed a pain point worth solving as our entry into that stretch. We found one:
15-page PDF called the crew list. Everyone needed it. Everyone hated using it.
Coordinators rebuilt and resent it by hand to 200+ people every time someone joined or dropped off. No version control. No single owner. Crew opened it, zoomed in, copied a number, and typed it into their own personal contacts, one at a time.
Every other piece of the production workflow had a tool. The crew list had nothing.
We assumed it was a simple as replacing the PDF with something digital
The obvious starting point: replace the PDF with a digital app that solved the pain points we just saw. Searchable. Department-organized. Call or text in one click. Built for how productions actually work.
Clear problem, tight enough to ship.
What we didn't know yet: we were only solving it for one user group.
Two iterations before we landed. Each one taught us what the next had to solve.
The first two versions were bottom-up: any crew member could bring the product in.
The third was top-down.
The Personal Contact Directory
Any crew member could upload the crew list PDF, and we turned it into a digital contact directory matched to Impact profiles. It was bottom-up and needed no coordination.
What worked. It kept work contacts apart from a personal address book. It also reused our existing search patterns, so it felt familiar and stayed small in scope.
What did not. Tapping call or message still sent people to iMessage, so they spent about 10 minutes a day in the app. Contacts from different productions mixed together. Coordinators were still building the PDF somewhere else.
The Active Production Directory
We scoped the directory to one production, with departments, favorites, and an active or archived state. Anyone could upload the crew list, and the upload invited everyone on it.
What worked. It fixed the mixed productions from Version 1. It also gave us a growth path, since any crew member could bring a whole production in.
What did not. Nobody owned the list. Old versions overwrote new ones, and we had to announce that only the coordinator uploads. There was still no messaging in the app.
Neither version gave us a clear answer, so we paused and went back to crew and coordinators. Across the three versions we had 15 conversations, and about five were with coordinators.
Crew liked both versions. They did not need either one. They spent about 10 minutes a day in the app, then went back to email, text, and calls.
The insight came while we were working on the hiring side of the product, which put us in front of coordinators. We had been solving for crew the whole time. Coordinators were still building the crew list somewhere else, and neither version had reduced their workload. They were also the way in, because a coordinator could convince a head of department to switch.
To move retention, we had to own communication and solve the coordinator workflow.
New Insights, New Direction
Communication On Platform
This means we had to invest in the communication framework. Not a workaround in iMessage. A purpose-built communication layer where every contact is verified, role-tagged, and scoped to the production.
Coordinator Owned Workflow
Every version so far had coordinators building the list somewhere else and uploading it in. That wasn't solving their problem, it was digitizing it. The list needed to live inside CrewChat from day one, owned by one coordinator, not maintained elsewhere and dropped in after the fact.
With the new north stars in place, the concept shifted entirely. Not a crew list feature — an on-set communication tool. Something crew & coordinators couldn't live without on a 14-hour shoot day.
Coordinator-Owned. On-Set Communication.
Version 3 went top-down. We were not going to convince 200 people one at a time, and an optional app does not get downloaded when the current tool is good enough.
One owner
A single coordinator is onboarded first and manages the crew list inside CrewChat: adding, removing, and updating. That ended the version conflicts from Version 2 and removed the coordinator's biggest chore.
An invite people act on
Invites go to the email addresses already on the crew list, so they arrive with the weight of the production office. We explored a QR code and cut it, because a scan was easy to skip.
Communication stays in the app
Calls, messages, announcements, and department threads all live in CrewChat. Nobody has to leave for iMessage.
Coordinator-Owned. On-Set Communication.
Version 3 committed to messaging as a core surface, not an afterthought.
Onboarding started with people, not a download link. Our CX team worked with one coordinator per production first, the same single point of ownership V2 had gotten right, and got them set up before anyone else touched the app. From there, invites went to the rest of the crew already organized by department, so the structure crew expected was there from their first login.
Once onboarded, coordinators could add or remove crew members inside the app itself, no more editing a PDF every time someone joined or dropped. V1 and V2 both left that part untouched, coordinators still maintained the list somewhere else and imported it in. Version 3 moved the list inside CrewChat for good, closing the one gap that had killed both earlier versions.
Standalone Or Integrated
One call remained: build CrewChat as its own app, or fold it into Impact's existing app. Standalone looked safer. It was simpler, faster to ship, and easier to kill.
Integration Won.
A standalone app would have solved the on-set problem and hidden Impact's hiring network from 200 crew using it every day. Get hired, work on set, find the next job: that loop only closes inside one app.
It was also cheaper than expected. Impact already needed network messaging, and the web version was responsive, so most of the work was frontend.
Tracking User Behaviors That Led To Evolution
We hadn't designed file sharing into CrewChat, but crew were using the announcements channel to distribute production documents anyway. The chat couldn't handle it: phones lagged, size limits were hit, no watermarking, no access controls. The live product told us exactly what to build.
File Distribution — the feature the announcements channel had become, done properly.
Coordinator-owned, watermarked per recipient, scoped access, built for the file sizes crew were actually sharing. It replaced the last piece of off-app workflow, email, and closed the full loop.
Proof, Not Projection
Ron Howard's EDEN was one of the first major productions to run fully on CrewChat.
Ground-level read on adoption from day one to wrap, not platform averages.
Time spent in Crewchat.
Studio deals, AMC, Skydance, Warner Bros, and Netflix, signed once CrewChat proved the platform could hold a production, not just staff one. Scenechronize was charging $30,000+ per production for less.
of new productions came in through referral, crew carrying the tool from job to job. That kind of adoption doesn't come from sales, it comes from people not wanting to go back.
"I don't know how I or any other production could do without it now. It set a new standard."
"CrewChat became my first choice for getting information from production instead of email."
"The most streamlined way to get messages out to everyone on set immediately."
Reflections
Cutting from MVP was the right call — every time
Call sheets and on-site status indicators were both designed and both cut. The live product validated every cut — shipping incomplete is a listening strategy.
Offline-first should have been a launch requirement
We knew signal was poor on soundstages and didn't prioritize aggressive caching. Given the constraint, offline-first should have been a launch requirement, not a retrofit.