I’d like to share some of my experiences using AppSheet to build business applications over the past few years.
From my perspective, AppSheet is a very interesting platform because it allows you to create business management applications without following the traditional software development process. For small and medium-sized businesses, it can be an effective way to digitize operations with relatively low cost and fast implementation.
So far, I’ve used AppSheet to build systems such as:
Project Management
Human Resources (HR)
Timesheets and Attendance Tracking
Payroll Management
CRM Systems
Task Management
Various Internal Business Workflows
One thing I’ve noticed recently is that Google seems to be giving less attention to AppSheet compared to a few years ago. This may simply be because AI is advancing so rapidly, and Google is investing heavily in products like AI Studio and the broader AI ecosystem.
That said, from a practical implementation standpoint, I still find AppSheet to be a powerful platform. When designed properly, it can deliver robust business applications without requiring a large development team.
In my experience, successful AppSheet projects depend heavily on good planning and optimization. Some lessons I’ve learned include:
Design the business process before building the app.
Minimize the use of Virtual Columns whenever possible.
Optimize formulas and data structures early.
Build a scalable database structure from the beginning.
Clearly define user roles, permissions, and workflows.
One area where I hope to see improvements is performance with large datasets. As applications grow and records increase, loading times can become slower than desired. Hopefully, future updates will continue to improve performance and scalability.
Overall, I still believe AppSheet is an excellent platform for building internal business applications, especially for small and medium-sized organizations that need solutions quickly without investing in full custom software development.
These are just my personal observations from working with the platform. I’d love to hear about other people’s experiences, best practices, and lessons learned from using AppSheet in real-world projects.
If you’re wondering what the screenshot above is, I actually built a custom tool using the AppSheet API to audit and analyze my applications.
The purpose of the tool is to extract and consolidate key information from an AppSheet app, including tables, columns, formulas, actions, bots, slices, views, and other configurations. This gives me a structured overview of the entire system in one place.
I primarily use it for:
Reviewing app architecture
Identifying optimization opportunities
Auditing formulas and business logic
Troubleshooting unexpected behavior
Documenting application structures
Checking for potential design issues before making major changes
One of the biggest benefits is that I can export the information and provide it to AI tools for further analysis. This helps me quickly review system design, identify bottlenecks, find optimization opportunities, and investigate issues that would otherwise take much longer to trace manually.
As AppSheet applications become larger and more complex, having a way to automatically inspect and document the app structure has become incredibly useful for maintenance and continuous improvement.
You have shared some interesting tips in the past about Google chrome extension to improve usability of the editor and HTML , AppSheet learning material etc.
Hope you have taken a look at the following announcement thread about ApPSheet’s upcoming product strategy. You may want to share your thoughts there about AppSheet possibilities.
Similar to you, I have created a lot of apps which are used internally for the SMB that I work at. Creating apps is not my day job/role but I saw a problem within the business and aimed to fix it.
I am good with Excel and the offline sync capabilities are invaluable for the type of work that we do, so I settled on AppSheet using Excel, then moved over to Google Sheets and 6 months ago moved over to Enterprise and a database.
Moving from Sheets to a DB has both been the most breaking and best thing that I have done.
Despite me checking the apps many many times before migrating I am still finding the odd column that is stored as a TINYINT so when I regenerate the table AppSheet says this is a true false column and it then goes unnoticed for ages and breaks a lot of things.
On the other hand, it so much faster than sheets. I don’t need to worry as much about tables getting out of hand in size and I am slowly moving a lot of bots and calculations out of AppSheet to the DB server. Last week I created utilisation charts based on time cards - I previously had this in Sheets but the only way to make it work was through the use of a lot of virtual columns which made the app fail to load some times. But now, the database does the heavy lifting and updates a view table twice a day in a fraction of a second which is more than enough for us.
The announcement the other week about the AppSheet team working less on new ideas and more on stability feels like the beginning of the end but if you work with the tools that AppSheet gives you and with some ideas from this forum you can get amazing results - I now only create PDFs using HTML and I am moving all charts to quick chart - as a business tool if you try to stay away from “mods” such as SVG making beautiful dashboards and try to keep it as simple as possible it is a great tool for data collection and spitting data back out.
I don’t know HTML, databases or how to re-create the utilisation charts that I made last week but I know how to collect and organise data and I know how to ask Gemini what I want and can spot errors that it is telling me, to do all of this.
By doing all of this I am now keeping my tables lean and producing deep insights in the data that is collected whist keeping everything reasonably fast - as a business tool you want it to work, be stable, be simple to navigate the UI and to be responsive, which is what AppSheet is (or can be become when backed with a database).
One thing I would suggest that could be helpful to others, is to also understand your database type in addition to its structure and design. At my business, we use one of a few different databases depending on the situation:
Google Sheet - Great for a quick process. Medium performance but can feed other use cases. No additional costs. Takes 3-10 seconds to save most changes.
BigQuery - Great for handling large data and building large reports (ie in Data Studio). Medium to slow performance, but fine when users are only making occasional edits. Costs for our size of use case is basically $0. Takes ~5 seconds to save almost every change (big or small).
Azure SQL Database - More upfront design effort, but handles data much quicker (ie an edit is saved in fractions of a second). Costs vary, but smaller DB options start around $15/month. Takes fractions of a second to save each change.
Therefore if I was asked to build an AppSheet for a team of 3-5 users to track projects they are working on where each person expects to open and use this once or twice a week, I would tend to choose BigQuery, as that is where we warehouse much of our organizations data anyways.
If I had an AppSheet where staff needed to review a long list of entries and assign charges to them, I would want a DB that can keep up with the speed that the user is making changes, and would make use of a SQL database for that.
The reality is that the work and effort of the product S curve which required most of the innovation, passion and investment drive ended with the Google purchase and the product is now on the end s-curve plateau maintenance and incremnetal development mode where it fulfils a niche and minor need within suite of Google apps.
Have you tried building an app using Cursor and next.js? My AppSheet journey is similar to your own, but my perspective on its value has radically changed since vibe coding has become so effective. I now spend far less time negotiating workarounds, I build apps that are far more contextualized, I spend zero time hunting through the interfaces to open my code. In short, I genuinely get work that is done better, faster and free of user licensing costs.
I found your information and thoughts on App sheet .to be so helpful and I am grateful for your input and love how you keep it organised so that it is very easy to follow neat and simple to make your work more manageable thank you so much for your response and I will be sure to be able to use this as well to make my work more efficient thank you.I love app sheet
AppSheet was NEVER anything but a tool for a niche need - if you can call quick internal business applications a “niche” need. I wouldn’t call it minor.
It works very well in that regard - especially when you consider “offline” mode / field employee capabilities and security.
But, I’m curious, since I haven’t reviewed Cursor yet (open to it)…how does this combination of development handle “offline” mode - which is critical to field oriented type businesses.
And if there is no licensing, how good is the secured access to the data?
These are the two critical areas I find hard for other platforms to beat!
Having spent the last six months exploring vibe-coding/agentic-development, I don’t think it’s a good approach for app developers without a broader software development background. It’s easy to create superficial apps with AI, but difficult to build robust, secure, efficient, maintainable apps. AppSheet provides a solid framework that handles much of that, allowing the app creator to focus on the business logic and presentation.
This structural integrity is exactly what makes AppSheet so valuable. Its robust foundation in relational data modeling provides the exact framework needed to ensure long-term maintainability and security.
By incremental I mean in terms of the developmental S-Curve where they have reached the upper plateau. By niche I mean in terms of the total Google portfolio.
AppSheets 2026 Product Strategy Update. Sounds as it on its way to the Google Graveyard. Its like Google Stadia all over again or Tables by A120. They pulling back on the updates. Then a year later then kill it.
I agree regarding the security issue. When working with other platforms or coding from scratch, managing security can be really challenging. Convincing clients to use a Google-backed platform gives them much more peace of mind about security.
Actually, I’m not that worried about it being discontinued. It’s already an integral part of the Google ecosystem, just like Looker Studio or Google Sites. They’ve been around for years without massive updates. It’s just that since it’s not Google’s core commercial product, new features are rolled out less frequently.
these are exactly the main elements why i love appsheet.
it is one of the best application i have used which works really well in offline mode and the user dont have to wait for the action to complete in the loading screen
