The Sales Capability Center — Part IV: Where Entries Fail, India Market Entry
A field guide for technology vendors selling into new market(India). Part IV of four
What we mean by a sales capability center
A sales capability center is the funded local unit a foreign technology vendor builds in order to sell in a market — people, judgment and relationships — as opposed to a global capability center, which is the unit a company builds in order to deliver from that market.
Part I was about who decides. Part II was about price. Part III was about the channel. This part is about how it ends when it ends badly, and what was visible long before anyone said so.
The short answer
Almost every failed entry was diagnosable at month six and admitted at month eighteen. The twelve months in between were spent watching the one number that moves last.
The mirror, stated plainly
Companies here are very experienced at running a new unit through its first year. They just do not apply that experience to selling units.
Global capability center, year one | Sales capability center, year one | |
|---|---|---|
Reviewed how often | Monthly, sometimes weekly | Quarterly, sometimes at the annual planning meeting |
Reviewed against | Hiring against plan, attrition, ramp time, quality, cost per seat | Revenue |
When a problem becomes visible | Early. Attrition and ramp move within weeks | Late. Revenue moves last, if at all |
What happens when a signal moves | Someone intervenes | It is explained |
Typical outcome of a bad year one | Corrected | Closed |
The same organisation that would never judge a new delivery unit on output alone will happily judge a new selling unit on signature value alone. Revenue is a lagging indicator in a market where the cycle is long. By the time it tells you something, you have spent the money.
Everything below is an attempt to give the selling unit the same early instrumentation the delivery unit already has.
Five post-mortems
These are the recurring shapes. Each is drawn from more than one entry.
1. The lonely seller
What happened. A well-regarded senior seller was hired as country lead. No presales, no marketing, no local support, reporting to a regional leader in another time zone who joined a call once a fortnight. Activity was high for two quarters. Nothing closed.
What was actually wrong. Part I: A deal here needs four or five people touched and a pilot resourced. One person cannot do that across a territory. The seller was not underperforming; the unit was understaffed by design.
Visible at. Month four — meetings held but no pilot started in any account.
What it cost. Lost time of upto 4 full time leads.
We lost valuable time and burned through budget before any meaningful traction in deal making -mentioned a senior account manager of a cybersecurity company.
2. The price that could not survive a spreadsheet
What happened. Strong technical evaluations, enthusiastic practitioners, several accounts reaching the formal comparison stage. Lost most of them at that stage, and was told the reason was budget.
What was actually wrong. Part II: The price was defensible only in a conversation. In a document, next to two other quotes, with no one there to explain, it lost. And the buyer had already priced the product in headcount equivalent, which the vendor never addressed because they never knew it was happening.
Visible at. The first lost bid — if anyone had asked what the comparison document looked like rather than accepting "budget" as the reason.
What it cost. 1.2 M USD
3. The partnership that was never refused
What happened. An agreement signed with a large services firm, announced, celebrated. Twelve months of cordial quarterly calls. No pipeline. No complaint from either side.
What was actually wrong. Part III: the vendor asked for demand creation and received a catalogue entry. No named accounts, no named individuals, no review consequence. The relationship existed entirely inside one enthusiastic person at the partner, who changed roles in month eight.
Visible at. Month three — no named individual at the partner had yet met a buyer alongside the vendor.
What it cost. 5.5 M USD — plus the year of direct selling not done while waiting.
4. The pilot that became the product
What happened. A large, prestigious first account agreed to a proof of value. It had no end date and no written definition of success. It ran for 12 months, absorbed most of the vendor's engineering attention, and expanded as the client found new things to test. It never converted; the client eventually restructured the project.
What was actually wrong. Part II: an unpaid, undefined pilot is not a step towards a purchase, it is free delivery. There was never a decision point because nobody had agreed what would trigger one, and no purchase order had ever been raised, so no one had ever had to find budget.
Visible at. The day it started, from the absence of a written success definition and an end date.
What it cost. 2.5 M USD — and the opportunity cost of every other account not worked.
5. The calendar that nobody read
What happened. A vendor built a plan on its own financial year, hired in [INSERT: month], and forecast first revenue within two quarters. Deals progressed well and then sat. The board reviewed at the twelve-month mark, saw no revenue, and cut the budget. Two of the stalled deals signed the following quarter, to a team that no longer existed.
What was actually wrong. Part I: budgets here are set before the financial year ends on 31 March, and uncommitted money is hard to find mid-year. The deals were never cold. The vendor's review point fell in the worst possible month.
Visible at. Before hiring, on a calendar.
What it cost. The market, and the deals.
The warning signs, by month
This is the table to put in front of whoever reviews the entry.
By month | If this has not happened, something is wrong |
|---|---|
3 | You can name a real buyer problem in your own words, heard from a practitioner rather than assumed |
4 | At least one pilot has started with a written definition of success and an end date |
6 | At least one account has a named person who owns budget, not just a person who likes you |
6 | You know what your price looks like in headcount equivalent, because a buyer told you |
9 | One purchase order has been raised for something, at any value |
9 | If you have a partner, a named individual there has met a buyer alongside you |
12 | One reference a prospect can ring personally, in this market |
18 | Your second deal came faster than your first |
The last row is the real test of the whole exercise. A first deal proves you can sell. A second deal that came faster proves you have built something. If deal two took as long as deal one, you have a lucky account, not a capability — and adding sellers will multiply the cost without multiplying the result.
The four sentences that should stop a review
When these are said, the correct response is a question, not a nod.
"The buyer went quiet." Buyers rarely go quiet for no reason. Either the sponsor lost the internal argument, the money was never attached, or an incumbent repositioned. Find out which. "Quiet" is not a status.
"We lost on budget." Sometimes true. More often it means the price lost a comparison you were not present for, or the value never reached whoever owned the money. Ask what the comparison document contained.
"It's in procurement." This describes a location, not a probability. Ask who inside procurement, against what timeline, and whether a purchase order has been requested.
"The partner is working on it." Ask which individual, in which account, and whether anyone from your side has met the buyer. If the answer is no to the last one, it is not pipeline.
The three mistakes underneath all of them
1. Applying home-market arithmetic to a different cycle. Seller ratios, ramp assumptions, forecast horizons and review points all carried over unchanged. Every one of them is wrong here, and they are wrong in the same direction: too fast.
2. Measuring the unit on the number that moves last. A delivery unit is measured on hiring, ramp and quality in year one. A selling unit gets measured on revenue and then closed before revenue was ever going to appear. The signals above exist to fix exactly this.
3. Withdrawing at the point of maximum sunk cost. The expensive years are one and two. Year three is when the references exist, the first accounts expand and the partner conversation finally works. Vendors who leave at month eighteen pay the entire cost of entry and collect none of the return — and they usually leave having built something that was about to work.
What survives, and why
The entries that worked, in the accounts we looked at, had little in common in product or size. They had four things in common in shape:
A unit, not a person. Seller, presales and marketing support, however small, funded together.
A local person with real authority. Someone who could price, commit and decide in the room. Buyers here read a seller who must phone home very quickly, and they price accordingly.
A review against signals, not revenue, for the first four quarters.
A three-year commitment made honestly at the start, rather than a one-year commitment renewed anxiously three times.
The last one is the hardest, because it has to be won internally before anyone is hired. The vendors who failed to win that argument were usually failing at month zero, whatever happened afterwards.
Questions a vendor should be asking
Which of the signals in the month table have we actually passed, and which have we been explaining?
Of our current pipeline — in how many accounts can we name the person who owns the money?
What did the comparison document look like in the last deal we lost, and did we ask?
Did our second deal come faster than our first, and if not, what did we conclude from that?
Is our review point falling in the worst month of the buyer's budget year?
Are we measuring this unit on something that moves before revenue does?
If we are at month eighteen and considering withdrawal — are we leaving before or after the expensive part?
Frequently asked questions
What is a sales capability center?
A sales capability center is the funded local unit a foreign technology vendor builds in order to sell in a market — people, judgment and relationships — as opposed to a global capability center, which is the unit a company builds in order to deliver from that market.
How long before we should expect first revenue?
Longer than the home market, mostly because of the pilot stage and the budget calendar. 6-12 MONTHS The more useful question is which signals should have moved by month six.
What is the most common reason entries fail?
Not product and not price. Understaffing the unit, then judging it on revenue before revenue could have arrived.
Is eighteen months long enough to know?
It is long enough to know whether the signals moved. It is usually not long enough to see the return, which is why so many withdrawals happen at the point of maximum cost and minimum reward.
What is the single best early indicator?
Whether the second deal came faster than the first. That is the difference between a lucky account and a working capability.
Should we cut our losses if year one produced nothing?
Look at the signal table before the revenue line. A year with pilots started, budget owners named and a purchase order raised is a year that worked. A year with meetings and enthusiasm and none of those is a year that did not, and adding sellers will not fix it.
We already have a delivery center here. Does that help us sell?
It helps with credibility, hiring and cost, and not much else. Selling and delivering are different capabilities, and the presence of one has not, in the entries we looked at, produced the other.
This completes the series. Part I: who decides. Part II: price. Part III: the channel. Part IV: where it fails.
How we know this: this piece draws on 126 interviews with technology leaders and enterprise sellers conducted between Jan to June 2026. Where a company is not named, it is because the interview was given on that basis.