Almost every request for an app starts with a manual process that eats up hours. Sometimes an app is the right solution. Usually the same problem can be solved with an automation, a form or a tool you already pay for, at a far lower cost and in far less time.
Start with the process that isn't working
When a business says it needs an app, it's usually describing a solution. First you need to understand the process: what isn't working, who does it today, how many times a day and what happens when it gets stuck. The answers determine whether you need development or a better-organized process.
The problems we run into repeat themselves: information moving between WhatsApp, Excel and email, approvals that depend on one person who is on vacation, and a report someone puts together by hand at the end of every month. None of them requires an app.
- Which process it is, and how many people it passes through
- How long it takes per week, in your own numbers
- Which error keeps recurring, and why it happens
- What information goes in and what comes out
- Who will maintain the solution 2 years from now
One more question: if this process stopped tomorrow, what would actually be affected. Sometimes it turns out nobody reads the monthly report someone prepares.
The cost of an app over five years
The development quote is only the beginning. An app keeps costing money after launch, so go through every line item before you decide.
| Item | How often it recurs |
|---|---|
| Specification and initial development | One-time, but it almost always grows along the way |
| Maintenance and bug fixes | Ongoing, for the life of the product |
| Developer accounts in the app stores | Annual subscription, separately for each platform |
| Updates for new operating system versions | At least once a year, whether you want it or not |
| Security, backup and server infrastructure | Ongoing, even with no new users |
| Replacing a developer or vendor | Every time someone leaves |
That's why it's worth calculating the cost over five years. An app that looks cheap in the quote can become the most expensive line in the digital budget, especially if it doesn't bring in new users.
There's also a cost that doesn't show up in the table: management time. Every product you own requires decisions, checks and people who remember how it's built. That's why it's worth revisiting the decision after a year.
When a custom app is justified
There are situations where there's no alternative to development. That happens when the need is unique to you, involves money or safety, and no off-the-shelf tool covers it.
- You sell the tool itself to customers, and it is the product
- The process needs hardware or a sensor with no existing interface
- Field work with no connectivity, syncing later
- High usage volume, where licensing an off-the-shelf tool costs more than building your own
- A regulatory requirement for full control over where the data lives
If one of these rows describes your situation, talk to a developer or a software house. We do not develop apps. We handle the infrastructure, the information security and the automation around the solution you choose.
Even in these cases it is worth splitting the work: develop only the part that has no alternative and leave the rest to existing tools. A small project goes live faster and is easier to maintain.
What you can solve with what you already pay for
Many businesses have a Microsoft 365 subscription they only partly use. It includes forms, shared lists, automatic approvals, organized storage and AI tools, and you can build a large part of the processes described here from them.
- A form that feeds a shared list, instead of messages scattered across WhatsApp
- An automation that sends an approval request to a manager and logs the answer
- A file that updates itself and generates the monthly report
- A reminder sent when a task passes its due date
- An AI assistant that drafts replies and summarizes customer inquiries
In the dashboard we build for every client you can see license utilization and AI tool usage, and cancel whatever is not in use. You can order the licensing itself through Software store of ours.
We implement this for clients as part of ongoing work, one process at a time, starting with whatever is causing the most trouble at that moment.
Map the process before you bring in tools
If the process is not in order, automation will only speed up the mistakes. So we start by mapping it on paper, in a one-hour session with the people who actually do the work.
- Write down the process step by step, as it is actually done today
- Mark every point where data is moved by hand from one tool to another
- Mark every point where the process stops and waits for one specific person
- Delete steps nobody can explain the reason for
- Pick one step to automate: easy to do and time-consuming
- Run it for two weeks alongside the manual process, then switch the manual one off
Document the result: one diagram and a few lines of explanation. Whoever handles the process a year from now, on your side or ours, will understand what runs and why.
Once the first automation has run for a month without intervention, move on to the next one.
Securing the data and measuring the savings
Before you connect a tool to company data, find out where the data sits, who at the vendor can see it, what happens to the data if you stop paying and whether the subscription is meant for business use.
- Sign-in through each employee's organizational account
- Two-factor authentication required, even on small side tools
- Role-based permissions only
- Automatic access removal on an employee's last day
- A clear answer on whether your data is used to train a model
Another question is what happens when the tool goes down. If a core process of yours depends on an external service, there should be a temporary manual procedure. Even a single page of emergency instructions helps.
Then you measure. Before rollout, record how much time the process took each week and how many errors it had. A month later, measure the same thing the same way. If the savings are smaller than the effort, turn the automation off. We run this cycle with clients as part of managed IT.
FAQ
How do you know if automation is enough instead of an app?
Ask whether the need breaks down into a form, a list, an approval and a report. If it does, automation is almost always enough. If the process requires special hardware or offline work, or it is the product you sell, development is justified.
Does MO-TECH develop apps?
No. We do managed IT, information security and cyber, AI consulting and implementation, and computing and mobile equipment at importer prices. If you need development, work with a developer, and we will take care of the infrastructure, permissions and security around it. The automations inside Microsoft 365 are also done by us.
How long does it take to implement one automation?
It depends mostly on people and less on the tool. What usually drags out a rollout is agreeing on how the process should work. That's why we start with a mapping session and only then configure the tool.
What do you do with an existing app that nobody maintains?
First, find out who holds the code, the app store accounts and the server. Then decide whether to hand off maintenance, freeze it or replace it with automation. Even an app that is no longer evolving needs security updates and backups, otherwise it becomes a risk.
More on this topic: Managed IT services · IT and license management dashboard · Licensing and software · IT and security assessment