How a Corporate IT Department Cut Vendor Sprawl by Half in One Quarter

The IT director at a 400-person Indonesian manufacturing corporate called us with a question that sounded simple. “How many software vendors do we have?” They didn’t actually know. The expense data lived in three accounting systems and the contracts lived in a shared drive nobody had organised since 2019.

This is the story of what we found, what we changed, and what they kept that we’d recommended killing.

The starting picture

After a two-week discovery, the actual number landed at 73 active vendors. Not 73 unused subscriptions — 73 contracts where money was leaving the building every month or year, allegedly in exchange for software the business needed.

The breakdown surprised everyone:

• 18 vendors were paying for things the business genuinely depended on. ERP, accounting, payroll, the obvious load-bearing systems.
• 22 vendors were paying for things one or two people used and the rest of the company didn’t know existed.
• 15 vendors were paying for things that two other vendors also did. Same job, three subscriptions.
• 9 vendors were paying for things nobody currently used. Shadow renewals. Free trials that became annual subscriptions because the credit card was on file.
• 9 vendors fell into “we think we use this but we’re not sure who”.

Total annual spend: Rp 4.2 miliar. The initial reaction in the room was disbelief. The director had estimated maybe Rp 1.8 miliar. The gap was almost entirely the second and third categories — small subscriptions that compound when nobody’s watching.

The intervention, in three phases

Phase 1 — Stop the bleeding (week 1–3)

We killed the easy ones first. The 9 unused vendors got cancelled within a week. That alone recovered Rp 380 juta a year. No business impact because nobody was using them.

The 22 single-user vendors got harder. Each one had a person attached to it who genuinely wanted the tool. The conversations were diplomatic. Of the 22, we kept 7 and consolidated 15 into existing tools the user didn’t know could do the same thing. Recovered another Rp 520 juta.

Phase 2 — Consolidate redundancy (week 4–8)

The 15 redundant vendors were the hardest because each had a stakeholder who’d championed it. Three different teams were paying for three different note-taking SaaS tools, two different chat platforms, two different file-sharing services.

The fix wasn’t picking the “best” one. It was picking the one most teams already used and migrating the others. We made one rule: if a tool had >50% adoption already, the other tools went away. That removed the politics of “but ours is better”.

This saved another Rp 680 juta a year and, more importantly, eliminated the cross-team friction of people on different tools trying to collaborate.

Phase 3 — Renegotiate the survivors (week 9–12)

For the 18 truly load-bearing vendors, we audited contracts. Five of them had auto-renewed at sticker price two years running. We renegotiated four of those with a credible threat to switch — saving Rp 290 juta a year.

We also enforced two new policies going forward:

• No new vendor without a written 90-day re-evaluation date. If it’s still being used at 90 days, it gets a real review. If it isn’t, it gets cancelled.
• All software spend flows through one approver. Not as a bottleneck — as visibility. The approver sees which department has 14 software subscriptions vs the one with two.

The end state

Vendor count dropped from 73 to 38. Annual spend dropped from Rp 4.2 miliar to Rp 2.3 miliar — a Rp 1.9 miliar swing in one quarter, or about a 45% reduction.

Less visibly but more importantly: the IT director now has a one-page dashboard showing every vendor, what it does, who owns the relationship, and when the contract renews. Three months ago that information didn’t exist anywhere.

What we kept that we’d flagged

Two cases worth mentioning. We’d recommended killing both. They overruled us.

The first was a project management tool used by exactly one team — a tool we thought could be replaced with the company’s existing platform. The team made a strong case that switching would cost them three months of productivity loss. They were right; we underestimated the embedded workflow knowledge. Sometimes the cheapest path is to leave a thing alone.

The second was a Rp 4 juta/month subscription nobody could explain. We were ready to kill it. Two weeks later, a finance manager mentioned in passing that it was the tool that produced their tax-compliance reports. Killing it would have cost them an external accountant’s time worth Rp 25 juta to recreate.

The lesson, which we now apply to every audit: just because nobody can explain a tool doesn’t mean nobody needs it. Always look for the silent dependency before cutting.

What this kind of engagement looks like

A vendor sprawl audit is one of the most consistently profitable engagements we run for corporates and larger SMEs. Cost is typically Rp 60–150 juta for a 12-week engagement, and the savings cover that within the first month for any company with more than 25 active vendors.

If you’ve never done this exercise — or you did it three years ago and nobody owns the discipline anymore — there’s almost certainly money sitting in your software spend that you could redeploy. An hour of conversation usually surfaces whether the math is interesting in your specific case. We do those at no cost.