How to keep your tech stack small on purpose
How to audit the software your business pays for, cut what nobody uses, and set a rule that stops the list growing again without starving the work.
Nobody decided to run nineteen subscriptions
Every tool on the list was a reasonable purchase on the day it was bought. Someone needed a scheduler. Someone trialled a design tool and the trial became a renewal. A departing employee left behind a project tracker that two people still open. Nothing here was a mistake and the total is still a problem.
The visible symptom is the card statement. The expensive symptom is that information about one customer now lives in four places and none of them agree. A new joiner needs a fortnight to learn where things are kept. Nobody can answer a straightforward question without opening three systems and the answer depends on which one they opened first.
What the subscriptions actually cost you
Consolidation does not automatically save money and you should not promise yourself that it will. Replacing six focused tools with one broad suite often costs more per seat and delivers weaker versions of the six jobs. The saving is rarely in the invoice. It is in the time nobody spends reconciling two systems and the mistakes nobody makes from reading the stale one.
Count the costs that never appear on an invoice. Every tool is an account that needs a password, a joiner to add, a leaver to remove and a security exposure if you forget. Every integration is something that can break silently. Every extra system is a place a staff member must remember to update. GST at 18% applies on service fees in India and foreign software billed in dollars carries exchange movement you did not budget for.
Who should cut and who should not
Cut if you cannot name the owner and the purpose of each tool in a single sentence. Cut if two products hold the same list. Cut hardest where the subscription renews annually and nobody noticed the last renewal.
Do not cut a specialist tool that is load-bearing for revenue in order to tidy the list. An ecommerce operation running a dedicated inventory product, or a studio on specialist editing software, loses real capability for a tidier spreadsheet. Do not consolidate into a single suite if the suite's version of your most important job is noticeably worse, because you will end up running the suite and the original tool and paying for both. And do not start this during your busiest quarter. Migrations fail when nobody has time to check them and a half-finished migration leaves you with more systems than you started with.
How to run the audit
Do not build the list from memory. Build it from the bank and card statements for the last twelve months, then from the app password manager, then from the list of applications connected to your email and storage accounts. That last source always reveals something nobody remembered authorising. Annual renewals hide from a one-month view which is why the window has to be a full year.
Then put each tool against a job rather than a department. For each row write the job, the owner, the annual cost, how many people opened it last month and what happens if it disappears on Friday. The rows that cannot be filled in are your answer. Adopt one rule going forward. One job, one tool, and a named owner before any purchase. Add a renewal calendar with a reminder two weeks before each annual date, because a renewal you see in advance is a decision and one you see afterwards is a refund request nobody will make.
Judging whether the stack is the right size
The metric that matters is how many systems a new employee must learn in their first week to do their job. If that number is above four the design is wrong regardless of what you are paying. A second useful figure is how many places a single customer record exists.
The metric that flatters is money saved in the first year. It is easy to produce by cancelling things nobody used and it tells you nothing about whether the remaining set works together. A business can cut its software bill substantially and still have three systems holding contradictory stock figures.
The common mistake
Buying an all-in-one platform to replace a sprawl and then running eight tools instead of seven. The suite arrives, two teams migrate, the third finds it cannot do the one thing they need, and their original tool stays. You now pay for the suite plus the holdout plus the integration joining them, and information that was in two places is in three. This is the most expensive way to tidy a tech stack and also the most common.
The quieter mistake is cancelling without migrating. A subscription is stopped to save money and the data inside it was never exported. Two months later somebody needs last year's figures and the account is beyond recovery. Export first, confirm the export opens, then cancel.
Keeping it small afterwards
Growth in the list is not caused by bad judgement. It is caused by nobody owning the list. Give one person the register, review it twice a year against the renewal calendar, and require that any new tool names the job it does and the tool it replaces. A stack stays small because somebody is accountable for its size rather than because everyone agreed to be disciplined.
This guide explains how something works. It is not legal, tax or financial advice, and rules change. Confirm anything regulatory with a qualified professional before you act on it.