Does your homepage explain what someone can buy from you? Say the product or service plainly, show enough to judge it, and make the next step easy to find.
My long-term goal is to design processes that can run again and again without depending on heroics. That means clear inputs, useful exceptions, and a reliable way to recover.
A customer complaint can point to a very specific screen that needs improvement. "I thought it was available" is worth tracing back to the page they saw.
What should your system do when it isn't sure? Ask you, hold the request, flag missing information? Decide that before it starts making confident guesses. #Automation
@DavidLinthicum A useful test: submit the same request twice, leave out a required field, and disconnect one dependency. Then see whether a staff member can understand what happened and recover without calling the developer.
@mashadrokova I'd love to see a small-shop example: one machine, one recurring bottleneck, and the actual setup and maintenance cost. That makes it much easier for an owner to judge whether automation fits.
@mashadrokova “What did you have to stop and chase down today?” seems like a useful starting question. Order updates, booking changes, missing information—that’s where I’d look for the first process to improve.
If a supplier changes a part number, the old number may still be on the customer's machine. Keep a traceable connection instead of making the old search disappear.
@Aiwithpoonam Before adding another subscription, map one job from start to finish. Where does information get entered twice? Where does someone wait for an answer? That gives you a shopping list based on actual work. Which combination has been easiest for a small team to maintain?
@iliasconfidency Replaying a real late shipment is a useful test. I'd also try missing tracking information, a partial delivery, and a customer who's already been contacted. Those cases reveal whether the workflow understands the situation well enough to act.
@productschool One question I like before building: “What will someone be able to finish tomorrow that they struggle with today?” It gives the developer a concrete outcome and the business a useful way to judge whether the change worked.
@eeelistar Ask the team: “What do you copy and paste every day, and what are you always chasing?” Those answers can reveal a useful first project. Pick one task, record how long it takes, and check whether the change actually helps.
@prukalpa A business's pricing rules, supplier relationships, and hard-earned exceptions should remain usable when it changes software. Can an owner export that knowledge in a form another system can actually understand? That's a practical test of ownership.
@alexalbert__ I'm building Ten Thirty Two around websites and workflows for small service businesses: repair requests, booking, and customer follow-up. I'd be interested in comparing how Claude handles those workflows, especially when an owner needs to review an action.
A dashboard can be crowded with accurate information and still leave the owner unsure what to do. Put decisions first when deciding what deserves space.
@CFIBNews A practical succession question: could someone else run the business for two weeks using the information already documented? Pricing rules, supplier contacts, open jobs, and customer promises are a good place to start.