A founder doing three-and-a-half crore in revenue, with a team of
nine, asked us in early 2026 how much he should be spending on
software. He had drifted, year on year, into a stack of monthly
subscriptions that totalled about thirty-two thousand rupees per
month, roughly one-point-one percent of revenue. He thought he was
under-spending. He also thought he was paying too much. Both could
be true, and in his case both were. The total was reasonable. The
allocation was not.
This essay is the allocation question. For a founder doing two to
five crore in revenue with six to twelve staff, what should the
software budget be, where should it go, and what to avoid spending
on. The numbers are specific because the question is specific. The
ranges are tighter than the consultant industry typically presents
them.
The total budget envelope
A defensible software budget for a founder at this revenue band is
one-point-two to two-point-five percent of annual revenue. For a
three-crore firm, that is three-point-six to seven-and-a-half lakh
per year, or thirty thousand to sixty-two thousand per month.
The lower end of the range is the cost of running a basic but
honest stack. Accounting (Tally), a few SaaS tools the firm cannot
operate without, hosting and domain, communications, and a small
allocation for ad spend tooling. The higher end is the stack that
supports active operational software and meaningful performance
marketing. The middle is where most firms sit comfortably.
Below one-point-two percent, the firm is starving the operation.
The team is paying for the gap in time and friction. Above
two-point-five percent, the firm is over-tooling. The team is using
a stack that exceeds what the operation needs to capture, and the
return on each marginal subscription is falling.
The exact figure inside the range depends on the firm's growth
trajectory and operational complexity. A firm growing thirty
percent year-on-year should be at the top of the range. A firm in a
steady state can sit at the bottom. A firm with high SKU complexity
or multi-channel sales needs the top of the range to handle the
volume. A firm with a single product line and a focused customer
base can hold the bottom.
The allocation that holds across firms
The split that holds, across the firms we work with at this scale,
is roughly the following.
Operational systems take about fifty percent. ERP, inventory, CRM,
order management, customer portal. This is the substrate the
business runs on day to day. The firm cannot under-invest here
without paying the cost in team time.
Marketing technology takes about twenty-five percent. The ad
platforms themselves are spend, not software cost, but the tooling
around them (analytics, attribution, landing pages, CRM
integration, retargeting infrastructure) sits inside the software
budget. For a firm with active digital acquisition, this allocation
is real and durable.
Accounting takes about fifteen percent. Tally remains, plus the
add-ons (TDS, GST filing tools, the CA's office's preferred
extensions). This is the smallest line item that produces the
largest perceived value because the firm cannot file returns
without it.
Miscellaneous takes about ten percent. HR and payroll software,
team communications (Slack, Google Workspace, Microsoft 365),
e-signature, shipping integrations, the occasional one-off tool.
This is the line that drifts upward without management. Founders
should review it twice a year.
For AKSD, the allocation looks
exactly like this. The operational layer, custom-built and
maintained under partnership, is the largest line. The marketing
stack, supporting active digital acquisition, is the second. Tally
sits at fifteen percent, supporting the CA's office on familiar
ground. The misc line is held tight and reviewed annually.
For Car Seat Wala, the allocation
skews slightly toward operations because the firm's inventory
complexity and customer customisation workflow demands a thicker
operational system. The marketing line is smaller because the firm
acquires through a different channel mix. The structure of the
allocation, however, is the same.
What to avoid spending on
The line items that consistently produce poor returns at this
revenue band are recognisable. A founder should not be paying for
them, and if they appear on the monthly subscription list, the
review should ask why.
The first is enterprise CRM subscriptions sized for a fifty-person
sales team when the actual team is four. The pricing tier carries
features the team does not use, and the implementation overhead
exceeds the value the team extracts. A focused CRM, or the CRM
embedded in the operational system, serves the actual workload.
The second is "all-in-one" platform subscriptions that promise to
replace several tools. The promise is rarely delivered. The
platform usually does one or two things well and the rest
adequately, and the firm ends up running the platform plus the
specific tools it failed to replace. The cost is paid twice.
The third is marketing automation tools sized for B2C scale when
the business is B2B with a hundred active customers. The
sophistication of the automation does not match the volume of the
list. A simpler tool, or the CRM's native automation, is
sufficient.
The fourth is collaboration and project-management tools that the
team installed during a sprint and never adopted. These show up on
the subscription list and produce no operational value. The fix is
a quarterly review and a hard cancellation of anything unused for
sixty days.
The fifth is AI subscriptions that have not been mapped to a
specific workflow. The recent appetite for adding AI tools is
strong. Most of them add no measurable productivity at the team's
current scale, because the team has not redesigned a workflow
around them. The right path is to add the AI tool only when a
specific workflow has been identified and the tool's value
hypothesis is testable. The longer treatment of this question is
in the AI essay.
When the build replaces the subscription
For founders past the three-crore mark, the question of whether to
build replaces some of the subscription line. The answer depends
on which subscription. The operational systems line is the most
common candidate. The firm has outgrown the off-the-shelf CRM or
the generic ERP, and the custom build delivers a system that fits
plus removes the subscription. The math takes the operational line
from a recurring monthly cost to a one-time build plus a
maintenance retainer, and the total over five years is usually
lower.
The accounting line should not be the candidate. Tally is honest
software at honest pricing. The build conversation should not
target the financial system of record. The longer treatment of
this distinction is in
the Tally diagnostic essay.
The marketing line is usually too small and too channel-specific
to be a build target. The firm rents the platforms; the firm
should also rent the tooling around them. The firm's investment
sits in the data layer that connects the marketing tooling to the
operational system, which is where decision infrastructure for
revenue lives.
The work we do under the internal systems
practice is structured for the founder who has decided the
operational line is the build candidate. The work under
performance marketing sits next to it,
linking the rented marketing platforms to the owned operational
substrate.
The annual review
The founder should sit with the subscription list once a year, in
the same way the founder sits with the vendor list. The question
to ask each line item is short. What is this for. Who in the team
uses it. What would break if we cancelled it tomorrow. If the
answer to the third question is "nothing", the line item should
go. The annual review usually returns ten to twenty percent of
the software spend to the firm with no operational impact.
The longer treatment of the build-versus-buy decision sits in
this essay. The
diagnostic for when accounting software is the bottleneck sits in
the Tally essay.
For a founder doing three crore in revenue, the software budget is
not the place to save money by under-investing. It is also not the
place to demonstrate seriousness by over-investing. The right
number is one-point-two to two-point-five percent, allocated in
the proportions above, reviewed annually, and treated as the cost
of the substrate the firm runs on.
If you want a working assessment of your current software stack
against the allocation above, Start a Conversation.