The short version
For most of my career, the people whose work decided my number did not report to me. Leading them anyway convinced me that the word "politics" gets the skill backwards.
- In enterprise selling, most of the people who decide your number sit outside your reporting line. At NTT that was every seller across 31 business units in three companies, and at AWS it was two of the three teams on my own account.
- What people dismiss as politics is a map of who decides, with what each person is paid on and what each needs to look good to their own boss. Working it in the open is the job.
- The plan does the work a reporting line usually does, provided every leader who runs to it can defend it in their own review.
- A cadence everyone can see keeps that plan alive, and the most useful question on it is where executives need to step in this week.
- Executive support has to be secured early. I learned that on a $42M bid we lost with the better offer, because I had assumed a partner's executives were behind us.
- In a consensual culture the decision is built before the meeting, and the quiet senior executive in the room can be the one who makes it.
1. Most of the people whose work decides your number do not report to you.
From the outside, enterprise selling looks like a seller and a buyer. Anyone who has carried a large number knows the seller is the visible end of a long chain: partner teams, technical and service teams, sellers in other business units, and the executives above every one of them. Very few of those people appear on the seller's org chart.
At NTT I led the Cross-Sell Program for the Americas, which spanned 31 business units in Dimension Data, NTT Communications and NTT DATA. None of those sellers reported to me. They worked for separate legal entities with separate profit-and-loss accounts, and the program itself reported into the NTT Global Business Office in Tokyo. Over its 22 months it produced more than $200M in contract value and more than $1B in pipeline.
Later, at AWS, I owned the account plan for the global MongoDB partnership and led the fourteen-person account team behind it, on a $172M number we grew 24.6% year over year in a down market. The sell-to team reported to me directly, and the partner and service teams aligned to my plan. Each of those two teams had its own leadership and its own remit.
Reporting lines ran in three directions, and the plan was the only thing all three teams shared. I think this is the normal condition of enterprise selling, and a leader who waits for authority before leading is waiting for something that will not arrive.
2. What gets dismissed as politics is a map of who decides and what each of them needs.
People call this skill politics, and they usually mean it as an insult. I understand why. Most of us have watched someone work an organization in private, trading favors and collecting credit, and the word stuck to that behavior for good reason.
Underneath the bad version sits something every leader needs: a map of who has the power to decide, with what each of those people is paid on and what each needs to look good to their own boss. A map like that turns ugly only when someone works it in secret. I put it on the table, and I build a plan that every person on it can see themselves in.
Allan Cohen and David Bradford made the same case decades ago in Influence Without Authority, first published by Wiley in 1990 and now in its third edition. Their model rests on what social scientists call the law of reciprocity: what you do for other people, good or bad, tends to be paid back over time. You get cooperation from people you cannot direct by working out what they value, which Cohen and Bradford call their currencies, and offering it. Their six steps line up closely with how I run an account.
| Their step | What it looks like in an account I run |
|---|---|
| Assume everyone is a potential ally | Partner and service leaders own named parts of the plan, with their own goals written into it |
| Clarify goals and priorities | One written account plan, reviewed against every team each quarter |
| Diagnose the world of the other person | The map: who decides, what each person is paid on, what makes each look good |
| Identify relevant currencies | Credit that lands where the deal books, and visibility with their own executives |
| Deal with relationships | A weekly whole-team call and a biweekly call with each team |
| Influence through give and take | The plan bends where it can, and trade-offs it cannot absorb go to the executives early, in writing |
The currency I lean on hardest is pay. At NTT that meant a cross-entity compensation model I designed, under which a seller kept full credit when a deal they found booked in a sister company. Without a model like that, asking a seller to bring a sister company into an account means asking them to give credit away. Once the pay works, the map gets much simpler, because helping you is already in everyone's interest.
3. The plan does the work a reporting line usually does, if each leader can defend it in their own review.
At AWS the account plan was the instrument. All three teams ran to it, and I reviewed every team against it each quarter. The partner and service leaders had no reason to take direction from me. What they needed was a plan they could stand behind when their own leadership asked what their team was doing and why.
That test changes how the plan gets written. One written for my own management chain describes my number. The version other teams will run to has to describe theirs as well, showing where their work appears and what they get credit for in front of their own executives.
A hypothetical: say the leader of a partner team you depend on has a different priority this quarter from the one in your account plan. Perhaps their organization has asked them to launch a new partner program, and your account is not where it starts. You cannot direct them, and escalating on the first day only teaches them to route around you.
I would start by finding out what that leader is measured on this quarter and who sets their goals, then draw the map in Exhibit 3. In this version the partner team leader is paid on partner-sourced pipeline and looks good when a new program lands. That points to where the plan can bend: make the account one of the first places the new program launches, with the pipeline credited to their team. If the two priorities still conflict after that, it becomes a conversation for the executives above both of us, held early and in the open, with the trade-off written down. Most of the time it never gets that far, because the plan bends enough for both teams to win.
4. A cadence everyone can see keeps the plan alive, and the weekly question is where executives are needed.
Left alone, a plan stops being the plan within a quarter. On the MongoDB partnership I ran three cadences against it. The whole team met weekly on what was pressing and where executives needed to step in. I held a call with each team every two weeks on its plans and operations, and once a quarter we held an account review against the plan.
Each cadence does a different job. The weekly call keeps everyone working from the same picture, and because the executive question is on the agenda every single week, asking for help becomes routine instead of a signal that something has gone wrong. A call with one team at a time gives that team room to raise its own issues without competing with the other two for the floor. At the quarterly review, everything is held to the written plan, which is where a leader who does not report to you can see that the plan they signed up for is the one being run.
Quarterly reviews were the heartbeat of the NTT program as well, at a larger scale, with operating-company leadership at every regional review. A cadence like that does something a reporting line cannot. It makes the plan public, so a leader who wants to change it has to do so in front of everyone else who runs to it.
5. Executive support is secured early, at the level where it holds, or it is not secured at all.
The $42M bid I describe in What an alliance actually runs on is the clearest lesson I have. Price was not the problem, because we had closed the gap to $15,000 and the customer's own review rated our offer better. We lost because I had assumed the partner's executives were behind us and never secured their commitment. Ever since, I have treated a partner's executive support as something to secure before the field needs it.
The NTT program was built the same way. Executives from each operating company signed off on it within its first 90 days, including who is prime, secondary and tertiary on a deal across all 31 business units, so the rule existed before the first conflict did.
Securing support early also changes what escalation means. When the executives above two teams have already agreed to the plan, an escalation asks them to settle a trade-off they saw coming, which is a much easier request than asking them to referee a fight. Exhibit 5 is the sequence I follow.
Escalating on day one feels decisive, and it costs more than it gains: the other leader learns that working with you means being overruled, and next time they route around you.
6. In a consensual culture the decision is built before the meeting, and the quiet senior executive may be the one who makes it.
My hardest lessons in this came from working across cultures. I worked inside a South African company and then a Japanese one, and the NTT program I led reported into Tokyo. That taught me to operate inside cultures that were not my own, down to reading the quiet senior executive in the room who is, in fact, the person who decides. At Dimension Data I negotiated Sony's global master services agreement in Japan.
Erin Meyer of INSEAD gives the clearest outside account of why the quiet executive matters. In The Culture Map and a 2017 Harvard Business Review article, she separates attitudes toward authority from the way decisions actually get made. Japan sits at the hierarchical end of the first scale and the consensual end of the second: people defer to the boss, and the decision is still built through broad agreement before anyone announces it.
The mechanics have names. Nemawashi, literally going around the roots, comes from the gardener's practice of preparing a tree's roots before moving it. In business it means the one-to-one conversations that build agreement, often before any formal proposal exists. Ringi is the formal half, where a written proposal circulates through the organization and each person who approves it stamps it with a personal seal. By the time the meeting happens, it usually confirms a decision that was built in private.
Seen that way, the quiet senior executive makes sense. Their view has already shaped the proposal, so they have no need to argue for it in the meeting. A seller who reads that silence as indifference, and pitches harder to the most talkative person present, has misread the room.
A second hypothetical makes the point. Say you are proposing a joint program to the leadership of a Japanese parent company. The meeting is cordial, the most senior person says almost nothing, and nobody objects or commits. A seller who expects decisions to happen in meetings walks out thinking it went well.
| What you see in the room | What it can mean | What I would do |
|---|---|---|
| The most senior person says little and asks one question late | They may be the decision-maker, and that question is the one that matters | Answer it completely, then follow up through the people who prepared the meeting |
| Nobody objects, and nobody commits | The groundwork has not been done | Go back to each stakeholder one at a time before asking again |
| A junior manager asks for detail you thought was settled | The proposal is being readied to circulate upward | Give them everything they need to carry it |
| The meeting ends with "we will study it" | A decision has not been built yet | Ask who else needs to see it, and offer to brief each of them |
I would read that meeting as a sign the groundwork had not been done. Before asking again, I would go back to each stakeholder one at a time, starting with the people who would prepare and circulate the proposal, and make sure the senior executive's view was known and built into it before anyone met again. It is the same rule as the $42M lesson, applied inside a different culture: secure the support early, at the level where it holds, before anyone has to give it in public.
7. AI keeps the map current and the cadence honest, and it cannot earn the trust that makes a leader bend.
Preparation is where AI has changed this work most, because sourced research on the organization behind a partner leader gets me to a first draft of the map before I need it. For the cadence, I would run the recordings of the weekly calls through AI into one consolidated read-out, so every team works from the same record of what was promised.
None of that earns the trust that makes a leader bend their plan for yours. That comes from a track record: a plan that served them last quarter, and an executive above both of you who had already said yes.
8. What this means for leaders who run matrixed teams
- Draw the map before you need it. For every leader your number depends on, write down what they decide and what they are paid on, then add what would make them look good to their own executive. Fill the gaps before the first conflict.
- Write the plan so other leaders can defend it. If a partner or service leader cannot show their own management where their team's work appears in your plan, they will treat it as yours alone.
- Settle the pay before you ask for help. Make sure helping you earns credit. A seller who is made whole wherever the deal books has every reason to bring you into their accounts.
- Put the cadence on the calendar and the executive question on the agenda. Ask every week where executives are needed, so escalation becomes early and routine.
- Secure executive support first, at the level where it holds. Never assume a partner's support. Get the yes above both teams before anyone in the field needs to ask for it.
- In another culture, do the groundwork before the meeting. Find the decision-maker, who may be the quietest person in the room, and make sure their view is in the proposal before it is presented.
Working inside a Japanese company taught me the version of this I use most. By the time leaders sit down together, the decision has usually been built one conversation at a time, before anyone walks in, and the leader who has done that groundwork is the one whose plan gets run.
Sources
- Allan R. Cohen and David L. Bradford, Influence Without Authority (Wiley, 1990, third edition 2017), on currencies, the law of reciprocity and the six-step model: Newswise, "Third Edition of Influence Without Authority Released", Babson College, November 2017.
- Erin Meyer, "Being the Boss in Brussels, Boston, and Beijing", Harvard Business Review, July-August 2017: hbr.org. Japan as hierarchical in authority and consensual in decision-making is also summarized by the Fund for Education Abroad and in an interview with Meyer published by HENNGE.
- Erin Meyer, The Culture Map (PublicAffairs, 2014): INSEAD, "New INSEAD Book Reveals Eight-Scale Framework".
- Nemawashi, its gardening origin and its role before formal meetings: GLOBIS Europe, "Nemawashi in Japanese Culture".
- Ringi, the circulated proposal and the personal seal, and how nemawashi supports it: Japan Intercultural Consulting, "How do the Japanese put in their two cents? Ringi!".
- NTT Cross-Sell Program results: NTT records, by reference. AWS MongoDB partnership results: AWS performance record, by reference.