ingridleiria Partnership Assessment
Partnerships fail from unclear economics and unowned execution far more often than from bad intent. The proposal arrives with a plausible strategic rationale, everyone agrees it makes sense, and nobody writes down who does the work, what each side actually earns in the case where volumes disappoint, or how it ends. Twelve months later there is a logo on a website, a quarter of a senior person's time consumed, and no revenue traceable to it.
The costs that hurt were avoidable on paper: exclusivity granted for a promise, which forecloses the better partner who appears eight months later; a channel arrangement competing with your own pipeline for accounts you were already working; and a dependency invisible until the renewal, when the other side knows exactly how much you need them. Behind most of it sits one habit, evaluating the partner's pitch rather than the partner. The deck they bring describes their upside case, sized with their assumptions, and it is not evidence.
When to use this, and when not to
Use it for any arrangement where two organisations commit their own resources to a shared outcome and share the result: an alliance, a reseller or channel agreement, a referral arrangement, a joint venture, co-marketing, a white label or embedded offering, or a corporate development opportunity short of acquisition. Use it equally for an existing partnership up for renewal that has not produced what was expected.
Do not use it to buy something. Where one side pays and the other delivers to a specification, that is a purchase, and vendor-evaluation applies with its weighted requirements, total cost of ownership, and reference calls. The distinguishing question is whether the counterparty is carrying risk and sharing upside, or invoicing.
Other adjacent cases. Selling an engagement to a client is proposal-writer, and the deck for that meeting is tailored-client-deck. Where the counterparty is an individual rather than an organisation, an associate or fractional leader placed onto engagements, this is not a partnership and the right skill is contractor-msa-and-task-order. Papering an agreed partnership, including referral fees, exclusivity and lead registration, is business-agreements-drafting, and formatting the recommendation for a decision meeting is decision-memo.
Commercial terms only. Legal structure, competition and antitrust questions, and regulatory approvals go to counsel.
What you need before starting
Your organisation's single objective, and the metric that would show it worked at twelve months. Distribution, capability, credibility, market access, cost sharing. Missing: this is the finding. A partnership under consideration because it was offered has no objective, and saying so is more useful than assessing it.
Evidence about the counterparty, gathered independently. Their strategy, their gaps, their recent moves, their other partnerships, their financial condition. From external-insights, public filings, their own announcements, and people who have worked with them. Missing: proceed, but mark every claim taken from their pitch as unverified and weight it accordingly.
The economics inputs. Volumes, conversion, pricing, margin, and what each side must invest in people and product work. Missing: build the model with visible assumptions and a sensitivity on the two that matter most, using financial-model-builder conventions. Never accept the partner's volume assumption as an input without a second source.
Who would own it internally, with hours. A named person and the time they can actually give. Missing: this is a stronger signal than any analysis. A partnership nobody has time to own will not be run, whatever the memo says.
Existing conflicts, and what the partner wants that is not money. Accounts already in your pipeline that the partner claims, channel conflict with direct sales, clients who compete with them, exclusivity already granted elsewhere; then exclusivity, data access, brand use, product commitments, roadmap influence. Missing: check the pipeline before modelling anything, since overlap converts revenue into margin given away, and ask directly about the non-monetary asks, which are where the real cost usually sits.
The method
State your objective and its twelve-month metric in one sentence each, before looking at the deal. Doing this after reading their proposal produces an objective shaped to fit what is on offer.
State their objective from evidence, not from their pitch. What do their recent moves, hires, filings and announcements suggest they need? Rule: where your evidenced view of their objective differs from what they have told you, that gap is the most important finding in the assessment. Then test compatibility: objectives can differ and still work, since credibility and distribution can both be satisfied at once, but they cannot both be ownership of the customer relationship.
Assess fit in four dimensions, each with evidence. Strategic: does this advance a priority you already have, or create a new one, which needs a stronger case. Customer: do their customers overlap with the ones you want, and does this conflict with existing channels or clients. Capability: what does each side bring that the other lacks, and what would it cost to buy or build it instead, stated as a figure. Operating and cultural: decision speed, quality standards, responsiveness, judged from the negotiation itself, which is the first joint project and the only unbiased sample you have. The test is countable, so count it: working days to answer a direct question, working days to return a document, and how many of their people had to approve each. The threshold: a median response slower than five working days while both sides are still courting each other should be assumed to double once the agreement is signed, and the operating model has to be priced for that slower speed or the deal declined. The disqualifier: a commitment made verbally and then contradicted or quietly dropped in the following document, twice. Once is a misunderstanding, twice is how the partnership will be run.
Model the economics for both sides under base, upside and downside, on incremental volume only. Subtract accounts already in your pipeline from the partner's claimed reach before counting any of it, and state the point at which the partnership beats the same resources spent on the next best alternative, usually direct selling. Rule: the downside case is the one that decides. If either side is unmotivated when volumes disappoint, the partnership stops being worked long before anyone terminates it, and the effort is lost anyway. Include the cost of the partnership itself: management time, enablement, marketing, and the deals given away.
Size the dependency at maturity. What share of revenue, capability, or market access would rest on this partner in three years, and what that dependency costs in bargaining power at renewal. A partnership that becomes a single point of failure has to earn a premium for it.
Design the operating model and the governance before agreeing anything. Named owner on each side with hours committed, what each delivers on what cadence, and a first ninety days with milestones. Then decision rights: what each side decides alone, what needs both, the escalation path with names, what data is shared and how often, brand and communication rules covering who may say what, and the mechanism for the conflicts you can already foresee, which are overlapping customers, pricing collisions, and competing products. Rule: if the partner cannot name their owner, the answer is not yet, whatever their enthusiasm.
List risks with their early signals, and separate contractual mitigations from operational ones. Partner underperformance, strategic drift, acquisition of the partner by a competitor, data or intellectual property leakage, exclusivity foreclosing better options. Rule: a risk with no observable early signal cannot be managed, only regretted.
Specify the exit before the entry. Term, renewal mechanics, termination for convenience and for cause, change of control, transition obligations, and what happens to shared customers, data, and intellectual property. A partnership with no clean exit is a merger without the price.
Write the recommendation with the deal points ranked and a walk-away. Proceed, proceed with conditions, pilot first, or decline. Where a pilot is right, define its scope, duration, success metric, and the decision at its end. Rank the negotiable points by value to you, and set the walk-away position before the first negotiation, not during it.
Worked example
Situation. Brightwell Data, a forty-five person analytics firm, was approached by Kestrel Integration, a mid-size systems integrator, proposing to resell Brightwell's analytics services into its client base at a twenty-five percent margin, with sector exclusivity for twenty-four months. Kestrel's deck claimed roughly forty clients in the sector. All figures in this example are US dollars.
Task. A recommendation for Brightwell's leadership within two weeks, with the deal points to negotiate and a walk-away, ahead of a second meeting Kestrel had already scheduled.
Action. The objective took two attempts. "Access to enterprise clients" was too vague to measure; it became reaching mid-market retailers that Brightwell's four-person sales team cannot cover, measured by four closed channel deals within twelve months, none from accounts already in the pipeline.
Independent evidence on Kestrel narrowed the claimed forty clients to eleven that plausibly fit, based on their published case studies and the sector filters that mattered.
The wrong turn was the first economic model, which treated all channel deals as incremental. A pipeline check found three of the eleven named accounts already in Brightwell's own pipeline, two at proposal stage, where the margin share would have been a pure loss on revenue Brightwell was going to earn anyway. The model was rebuilt around a deal registration rule, which became the second most valuable term in the negotiation.
Base case: six deals in year one at an average of 60,000 dollars, so 360,000 dollars of revenue, 270,000 dollars net of the margin share. Against that, the cost of running the partnership was 0.4 of a person at roughly 48,000 dollars, plus enablement. Downside case: two deals, 120,000 dollars of revenue, 90,000 dollars net, which does not cover the management cost plus the opportunity cost of the same time spent selling directly. That was the finding: fine if it worked, quietly negative if it half-worked, which is the likeliest outcome for a first channel arrangement.
Exclusivity was refused in its proposed form, because twenty-four months in exchange for an unevidenced volume estimate transfers all the risk to Brightwell. The counter was exclusivity earned rather than granted: none at signature, twelve months once four deals had closed.
Result. The recommendation was proceed with conditions, structured as a nine-month pilot with three named accounts, deal registration on every opportunity before any margin share applied, a named partnership owner on each side with four hours a week committed, monthly pipeline review, and no exclusivity until the threshold.
At nine months, three deals had closed rather than the four in the target. The partnership was extended for six months rather than converted, on the basis that two of the three had come from the same Kestrel salesperson, which is a person-level dependency rather than an organisational channel. That distinction was only visible because the reporting had been specified at the start.
A second scenario, where it goes differently
The same firm was offered a partnership by a trade association: co-branded research, a speaking slot at their annual conference, and use of their member logo, with no revenue share in either direction.
The economics test barely applies here, and forcing a revenue model onto it produces a fiction. What replaces it is an attention budget: the partnership would consume roughly fifteen days of a senior person's year, and the metric became qualified conversations with member organisations rather than revenue. Exit mattered less, because there was little to unwind. Brand rules mattered much more, because the whole value was reputational and the association's approval process for joint material was slow enough to affect the publication schedule.
What did not change: the named owner with hours, the evidenced view of what the association wanted (membership renewals, for which credible research is the lever), and the early signal that would show it was not working, which was defined as fewer than six member conversations by month six.
Output
An assessment memo, one to two pages plus the model as an appendix.
PARTNERSHIP ASSESSMENT: [counterparty]
Recommendation [proceed / proceed with conditions / pilot / decline] + the strongest reason
Our objective [one sentence] + [12-month metric]
Their objective [one sentence, from evidence] + [source]
Compatibility [where the objectives align and where they diverge]
FIT
Strategic / Customer / Capability / Operating, one line each with its evidence
ECONOMICS
| Case | Volume | Revenue to us | Cost of partnership | Net | Motivated? (both sides) |
| Base | | | | | |
| Upside | | | | | |
| Downside | | | | | |
Overlap with existing pipeline: [accounts, value]
Payback against the next best use of the same resources: [when]
Dependency at maturity: [share of revenue or capability, and what it costs in bargaining power]
OPERATING MODEL
Owner (us) [name, hours] Owner (them) [name, hours]
First 90 days [milestones] Reporting [what, how often]
Decision rights [alone / joint / escalation] Conflict handling [mechanism]
RISKS
| Risk | Early signal | Mitigation | Contractual or operational |
EXIT
Term, renewal, termination for convenience and cause, change of control,
shared customers, data and IP on exit
DEAL POINTS, RANKED
1. [point] target [x] walk-away [y]
2. ...
Decision requested by [date]
Failure modes
Assessing their pitch instead of the partner. Recognise it because every number in the model traces to their deck. Rebuild the volume assumption from your own evidence, and mark anything that cannot be independently checked.
Upside-only economics. Recognise it because the downside case is absent or is the base case with a smaller number. Model the case where volumes disappoint and check that both sides are still motivated to work it; if either is not, the partnership dies quietly rather than formally.
No named owner, or an owner with no hours. Recognise it when the memo says the partnership will be owned by a function rather than a person. Ask for a name and a weekly hour count on both sides, and treat refusal as the answer.
Exclusivity granted for a promise. Recognise it when exclusivity appears at signature and the volume commitment does not. Make exclusivity earned against a threshold, with a term short enough that a mistake is recoverable.
The partnership standing in for a sales motion that does not work. Recognise it when the objective is growth in a segment where your own direct effort has failed and nobody can say why. A channel amplifies a working motion and does not create one. The related error is announcing before the operating model exists: the press release removes the pressure that would otherwise have produced the plan.
Edge cases
The partner is much larger than you. Assume asymmetric attention: their side will be one of many and yours one of few. Ask which of their people carries it, how many partnerships that person holds, and where you sit in their priorities, and set exit and change of control terms tightly, because their strategy will change without reference to you.
They compete with you in one segment. Do not assume goodwill will manage it. Define the boundary by named accounts or named products, write the conflict mechanism, and limit what data crosses; where the boundary cannot be drawn cleanly, that is a decline.
An existing partnership that has underperformed. Assess it as though it were new, then add the question the fresh case does not have: what specifically was tried, by whom, with how many hours. Most underperforming partnerships were never actually worked, and reviving one is cheaper than starting another, but only when the reason is effort rather than fit.
The counterparty is an individual. A fractional executive or an associate offering to bring work is not a partnership even when both sides use the word. Use contractor-msa-and-task-order, which puts the master agreement in place before any engagement-specific paper.
Equity is wanted or offered, or the idea came from your own chief executive. For equity, stop the commercial assessment and involve counsel and the board, since instruments, valuation and control sit with business-agreements-drafting and legal review. For an idea brought back from a conference by the person who will decide, the analysis is unchanged but the delivery is not: lead with what would have to be true, and rehearse it with principal-simulator, because a recommendation to decline is more likely to be read when it arrives as conditions rather than as a verdict.
Quality bar
- Both sides' objectives are stated, and the counterparty's is evidenced independently of their pitch.
- The economics show base, upside and downside, and confirm both sides stay motivated in the downside.
- Overlap with the existing pipeline is quantified and removed from the incremental case.
- A named owner with committed hours exists on each side, with a first ninety days.
- Every risk has an early signal you could actually observe.
- Exit terms cover termination, change of control, and what happens to shared customers, data and intellectual property.
- Deal points are ranked by value with a walk-away position set before negotiation.
- Where the honest conclusion is that you would be the junior partner absorbing the work, the memo says so plainly.
Adapting this to your context
This comes from professional services and software firms of twenty to three hundred people evaluating channel, referral and alliance proposals from larger counterparties. The tests hold; the settings need resetting.
- The five working day responsiveness threshold. Regulated counterparties, universities, hospital systems and public bodies run on committee calendars and are slower without it meaning anything. Set the threshold from how fast that sector moves when motivated.
- Exclusivity earned rather than granted. The example uses four closed deals in twelve months. Make it roughly one sales cycle's worth of proof, so a nine-month enterprise cycle needs an eighteen-month test rather than a twelve-month one.
- The nine-month pilot and four hours a week. Shorten the pilot to a quarter for a transactional product, and lengthen it past a year where a procurement round is the unit of evidence. A partner much larger than you needs more hours.
- The economics test. Where no money changes hands, as with the trade association, replace revenue with an attention budget in days and a non-financial metric, and keep everything else.
- What not to change. The counterparty's objective is evidenced independently of their pitch, and the downside case decides. If either side stops being motivated when volumes disappoint, the partnership is already over.
Related skills
external-insights gathers the independent evidence this assessment depends on, and market-research sizes the opportunity the partnership is meant to reach. financial-model-builder builds the economics and expected-revenue-estimation sizes the channel before it has a track record. decision-memo is the format the recommendation is delivered in, and structured-problem-solving structures the analysis behind it. vendor-evaluation is the right skill when the counterparty is paid to deliver rather than sharing risk and upside. business-agreements-drafting papers the partnership, referral or reseller agreement once terms are agreed, and contractor-msa-and-task-order applies when the counterparty is an individual joining engagements. proposal-writer and tailored-client-deck cover the client relationship, which is a different transaction with a different test, and program-management runs the first ninety days once a partnership is approved.