Segmenting and Tiering Your Partner Ecosystem

Segmentation is not tiering with a softer name. It is the decision that has to happen first: what kind of partner is this, judged by what they are, not by how well they are currently doing. Business model, reseller, referral source, systems integrator, managed service provider, independent software vendor. Route to market. Vertical specialism. Geography, where that changes the buying motion rather than just the postcode. None of that says anything yet about performance. It says what game this partner is even playing, which is the same distinction I drew when I argued that partner types are business models, not roles.

Get this step wrong, or skip it, and tiering becomes an exercise in comparing incompatible things on one scale. A global systems integrator and a three-person compliance boutique were never going to produce the same revenue chart, not because one is better than the other, but because they are not the same kind of business, with different capacity, different deal sizes, and a different route to the same customer. Judge them on one universal ladder and the only thing the chart tells you is which partner happens to be larger, which is precisely the information you already had before you built the chart.

Done properly, segmentation produces a small number of groups, each with a tiering ladder that actually fits the shape of business inside it. A ladder built for systems integrators should weight executive sponsorship and committed delivery capacity heavily, because that is what separates a serious relationship from a logo. A ladder built for boutique specialists should weight sourced pipeline per practitioner, because headcount is the scarce resource in that segment, not brand reach. Only once a partner's segment is settled does the tiering question, what have they contributed, what do they realistically hold in potential, start to mean anything at all.

I have been fortunate enough to onboard a handful of large consultancies over the years, and the tiering question never arrived as a tidy strategy debate. It arrived the moment someone asked which category a newly signed firm should go into on our public partner directory. No deals had closed yet. No pipeline existed. And the real question sitting underneath that categorisation dropdown was not analytical at all: are we about to upset one of the largest consultancy businesses in the world by not putting them in our highest category?

That is the version of this problem I hear most often, and it rarely stays theoretical for long, because the partner can usually see exactly where you have placed them. Assume the segmentation question is already answered, the firm is squarely a consultancy or systems integrator, no debate there. Where do they sit on that segment's ladder?

Put them in the top tier and you are rewarding the signature, not the contribution, which is the exact trap I flagged when I wrote about incentive design: tiers only mean something if they reflect value created, not the size of the name attached to them. Put them in the base tier by default, alongside every other unproven partner, and you risk starving a genuinely promising relationship of the enablement and attention it needs to ever produce anything, then pointing at the empty pipeline eighteen months later as proof the brand did not matter after all. Neither answer is a decision. Both are a shrug dressed up as governance.

The instinct to default to the top tier is brand doing the talking in place of evidence, reputation standing in for revenue. The instinct to default to the base tier is subtler and, in my experience, more common, because it feels prudent. It is not prudence. It is simply refusing to make a call and calling that discipline.

The reason tiering conversations go in circles is that tiering is quietly answering two separate questions at once, and most models only ask one of them.

The first question is backward-looking: what has this partner actually contributed? This is where the four metrics I laid out when writing about measuring partner productivity properly belong: sourced pipeline, partner-attached win rate and cycle time, expansion in partner-managed accounts, and influenced revenue. A tier built purely on these is honest, but it is also blind to anything that has not happened yet, which is precisely the gap a newly signed partner falls into.

The second question is forward-looking: how much potential does this relationship realistically hold? This is closer to the weighted scoring I use to decide which partners to pursue in the first place, customer overlap, business alignment, strategic timing, and the rest, except it is being asked about a partner you have already signed rather than one you are still evaluating.

Most tiering models pick one of these questions and ignore the other. A pure revenue tier cannot place a new partner anywhere sensible. A pure potential tier never demotes anyone, because everyone still has potential long after the evidence says otherwise. You need both axes running at once, and you need to be explicit about which one is doing the talking at any given moment.

The practical fix is a tier that is neither the top nor the bottom: a provisional placement, time-boxed and reviewed on a fixed cadence, for any partner who does not yet have a contribution record. Call it what you like internally, developing, emerging, incubating; the label matters less than the discipline behind it.

This is not a hypothetical fix invented for a Monday morning problem. AWS runs exactly this shape at a scale most of us will never manage directly: new partners join as Registered, a status held apart from the graded Select, Advanced and Premier tiers above it, and only move onto that ladder once there is enough evidence to place them on it properly. Not badged as emerging, in their case, but functionally the same instinct, and worth noting because a partner network that size did not add a fourth status by accident. It added it because neither of the shortcuts, straight onto the bottom rung of the real ladder, or waved onto the top of it, survives at volume.

A partner enters this tier on potential alone, scored the same way you would score a prospective partner before signing. What differs from a permanent bottom tier is that it comes with two things a bottom tier should not: a defined level of enablement investment proportional to the potential score, and a hard review point, ninety days, two quarters, whatever suits your cycle, where the contribution metrics either confirm the potential or they do not. A high-potential score buys a partner a fair chance to prove itself. It should never buy a permanent seat at the top table, and it should never be confused with an excuse to ignore the partner until the review comes around.

This is also where the review has to be honest in both directions. A partner that clears the bar graduates on evidence, not on the strength of the logo that got them the provisional placement in the first place. One that does not is either given a second, narrower window with a specific reason to believe the picture will change, or moved down, regardless of how the original signing announcement read on LinkedIn.

None of this needs an elaborate stack, and the building blocks I described when writing about scaling from ten partners to fifty apply just as directly here: the discipline has to exist before the tool does, and the tool's only job is to carry that discipline without adding weight of its own. Introw and Kiflo both do this well for a lean team, pairing simple partner-type grouping with a tiering ladder a partner manager can explain on a call rather than a binder only an administrator understands, and Introw's own guide on structuring tiers makes exactly that case: three levels, not five or seven, trail markers up a hill rather than a hedge maze. PartnerStack suits a different segment of the same problem, high-volume, SaaS-native, referral and resale heavy, where tiering works as a rules-based unlock rather than a judgement call. Impartner is the one built for genuine scale, carrying the same logic across a partner base large enough that nobody could track it by feel, with an audit trail that earns its keep once the numbers get big. None of the four will tell you what should have earned a partner into a tier in the first place. What a simple model, well chosen, buys you is the thing that broke last time the numbers grew: a system that still holds at fifty partners instead of quietly collapsing back into someone's memory.

A segment is not a judgement and a tier is not a compliment. One tells you what kind of partner you are looking at. The other tells you how well they are playing the game that segment demands, and until the evidence exists for a new arrival, the honest answer is a provisional seat with a date attached, not a guess dressed up as either generosity or caution.