June Offer Every MAX plan gets a fully custom-built system Free custom system worth $1,500-$10,000 · worth $1,500-$10,000

Referral partnerships, with numbers you can defend

It finds the businesses near you worth approaching, explains why each pairing works, and writes the proposal. Every figure is marked verified or estimated, and the ones that cannot be substantiated are left blank.

Which is the difference between a partnership and a polite no.

Get Started

Every number says where it came from

Ask any AI to write you a partnership proposal and it will produce something confident and readable containing a market size, a customer overlap percentage and a revenue projection. None of those numbers are real. They are written in the same tone as the true parts, which is what makes them dangerous: you cannot tell them apart, and neither can you a week later when you are in the meeting.

So every quantitative field in a proposal here carries its origin, and that origin is visible on the document rather than in a footnote nobody reads.

Verified

Figures from your own CRM, given to you directly by the client, taken from a verified Google Business Profile, drawn from a partnership you have actually run before, or computed from those. These are the lines you can stand behind when somebody asks where the number came from.

A proposal that is mostly this is a strong one, and the readiness score says so.

Estimate

Anything from an industry benchmark or generated by the model. Marked as an estimate on the face of the document, carrying a confidence level, a range where one makes sense, and the assumptions it rests on written out.

These are editable, and you should edit them. You know your business better than a benchmark does, and an estimate you have corrected becomes a number you can talk about.

And sometimes: blank

The two that most often come back empty are market size in currency and the audience overlap between the two businesses. Both are unknowable without data neither side has handed over, and both are exactly what a generated proposal normally invents.

Leaving them out is enforced behaviour rather than a setting. It looks worse on the page and it is the single most valuable thing this does, because a business owner who checks one invented figure stops reading and you do not get a second attempt.

What it actually produces

  • A list of real named businesses near you worth approaching, with a reason for each pairing
  • Twenty partnership shapes, from a two-way referral to an event sponsorship
  • A proposal document where every figure is labelled verified or estimated
  • Estimates you can edit, with their assumptions and confidence shown
  • Unsupportable figures left blank rather than invented
  • A readiness score with a specific list of what is missing and how to fix it
  • A send gate that stops a thin proposal going out on impulse
  • The outreach message, an agreement outline and an email version, all from the same document
  • Status tracking through sent, viewed, accepted, rejected or revision requested
  • Scoped per client and workspace, with your branding rather than ours

Not included: commission tracking, payouts, partner portals and affiliate attribution. That is a different product and the comparison further down says so plainly.

Finding who is worth asking

The reason most businesses have no referral partners is not that they disagree with the idea. It is that sitting down and working out which twenty local businesses to approach, and why each one would care, is a day of work nobody has. So it never starts.

Adjacent, not competing

The same sweep that maps your competitors also surfaces the businesses serving your customer at a different moment. A bridal shop and a hair and makeup studio. A car dealership and a mortgage broker. A car wash and a garage. Each suggestion comes with the reason stated, because the reason is what you will actually put in the first line of the email.

Real businesses, not categories

Named businesses with a website, a rating and a review count, discovered by actually sweeping your area rather than by asking a model what kinds of business might be nearby. That distinction is the whole difference between a list you can work through on Monday and a list you have to go and research first.

The same run also surfaces what your competitors are doing, plus the chambers, networks and directories worth being listed in.

Or somebody already in your pipeline

A proposal can be generated against an existing deal or contact rather than a discovered business, which is the case people underestimate: the best first partner is frequently a customer who already likes you, or a deal that went quiet for reasons that had nothing to do with whether they rate you.

Twenty shapes a partnership can take

The shape changes the whole document, not just a label at the top. Picking the right one matters more than the wording.

Partnership types and when to use each
ShapeTypesWhen it fits
Sending each other businessTwo-way referral, one-way referral, revenue share, commission partnershipThe default, and the one most local partnerships actually are. Two-way is where it starts unless one side clearly has more to give.
Selling togetherCross-promotion, co-branded offer, service bundleWhen the two businesses serve the same customer at different moments and a combined offer is genuinely better than either alone.
Offering something to their peopleCustomer discount, loyalty benefit, employee benefitThe easiest first partnership to get agreed, because one side is only being asked to pass on a perk. A good way in with a business that does not know you.
EventsEvent sponsorship, event vendor, event exhibitorWhere the ask is specific and time-boxed, which makes it easier to say yes to than an open-ended arrangement.
Audience and contentContent collaboration, newsletter exchange, social collaboration, local SEO collaborationWhen neither side has budget but both have an audience. The local SEO one is underrated: two businesses linking to each other properly is worth more than most link building.
CommunityCharity partnership, community partnership, customSlower to pay off and better for reputation than pipeline. Worth being honest with yourself about which one you are doing.

It will tell you when it is not good enough to send

Every proposal gets scored strong, fair or thin, with a list of the specific things weakening it and how to fix each one. Thin almost always means the same thing: the document is mostly estimates, because nothing real has been connected to it yet.

Below the threshold, sending is refused. You can override it, but you have to type a reason first. That is deliberately a speed bump rather than a lock, because the failure it is guarding against is not somebody being reckless, it is somebody in a hurry firing off a half-built document to a business owner they get exactly one attempt at.

Most software would treat a warning as sufficient. The reason this one blocks is that partnership outreach is unusually unforgiving: a weak proposal does not get a reply asking for more detail, it gets silence, and you have burned the introduction.

Against a partner platform, an AI, and a template

There is no direct competitor to this, so the comparison is against what people actually do instead. One of the three is a real product doing a different job, and the table says so rather than pretending we beat it.

Partnership finding and proposal capability by approach
CapabilityInflowavePartnerStackChatGPTA template
Finds businesses near you worth partnering withYesNoPartlyNo
Says why a specific pairing makes senseYesNoPartlyNo
Writes the proposal documentYesNoYesPartly
Labels every number as verified or estimatedYesNoNoNo
Leaves a number blank rather than inventing oneYesNoNoNo
Scores whether the proposal is strong enough to sendYesNoNoNo
Blocks you from sending a thin oneYesNoNoNo
Produces the outreach message and an agreement outline tooYesPartlyYesNo
Tracks the proposal through to accepted or rejectedYesYesNoNo
Runs an affiliate programme with payouts and commissionsNoYesNoNo
  • Finds businesses near you worth partnering with: PartnerStack manages partners you already have; finding them is not what it is for. An AI can list plausible categories but cannot tell you the names of businesses actually operating in your postcode.
  • Says why a specific pairing makes sense: Adjacency with a stated reason, e.g. a bridal shop and a hair and makeup studio, or a car dealership and a mortgage broker. An AI produces the same reasoning generically; the difference is that ours is attached to a real named business with a website and a review count.
  • Writes the proposal document: An AI writes a better-reading first draft than most people would. This row is not where the argument is.
  • Labels every number as verified or estimated: The row that matters. Each figure carries where it came from, from your own CRM, from a verified Google Business Profile, from a past partnership, computed, or an AI estimate. An AI writer presents a guess and a fact in identical prose.
  • Leaves a number blank rather than inventing one: Market size in dollars and audience-overlap percentages come back empty when the data cannot support them. That is enforced behaviour, not a default. It is also the single most common thing an AI-written proposal fabricates.
  • Blocks you from sending a thin one: A proposal below the readiness threshold is refused. You can override, but you have to type a reason, which is a deliberate speed bump rather than a lock.
  • Produces the outreach message and an agreement outline too: The proposal, the email that sends it, and the outline of the agreement all derive from one document, so the three cannot drift apart. An AI will happily write all three and contradict itself between them.
  • Runs an affiliate programme with payouts and commissions: We do not do this at all. PartnerStack and Impact are built for it. If what you need is to pay a network of affiliates, that is their product and not ours.

Where the others are stronger. PartnerStack and Impact.com are proper partner-management platforms with payouts, commission tiers, partner portals and attribution. We have none of that, and if your problem is running an affiliate programme at scale then you want one of them and this page is not relevant to you. An AI writer will also produce a smoother-reading first draft than our generator does, because it is unencumbered by having to be true. And a Word template costs nothing, which is a real argument when you are sending two proposals a year.

The gap this fills is the bit before a partner programme exists: you have no partners yet, you do not know which local businesses are worth approaching, and writing a proposal to each one by hand is why it never happens. So it finds the businesses, explains the adjacency, and produces a document where every number is labelled with where it came from and the ones that cannot be substantiated are left blank. That last part is the difference between a proposal that survives being read carefully and one that gets you a polite no.

Competitor capabilities from vendor documentation, checked 30 July 2026. Verify any row that decides your choice.

One at a time

Each starts with the case for the alternative, and on the first one that case is simply that it is a different product.

Inflowave vs PartnerStack

Use theirs when: A completely different job, done well. If you already have partners and need to run a programme with commission tiers, payouts, attribution and a partner portal, PartnerStack does all of that and we do none of it. Nothing on this page is an argument against buying it.

Use ours when: The problem before that one. PartnerStack does not find you a partner and does not write the pitch that gets a local business to say yes. If you have no partners yet, a partner-management platform manages nothing.

Inflowave vs writing it with ChatGPT

Use theirs when: Free, fast, and honestly a better writer. For a proposal to somebody you already know well, where the relationship carries it and the document is a formality, this is the sensible option and we would use it too.

Use ours when: It cannot know which businesses near you are worth approaching, and it will confidently invent a market size, a customer overlap and a revenue projection, presented in exactly the same tone as the true parts. Send that to a business owner who checks one number and you have lost the partnership and the credibility. Ours marks its estimates as estimates and leaves the unsupportable ones empty.

Inflowave vs a proposal template

Use theirs when: Free, and if you are sending two partnership proposals a year the effort of anything else is not worth it. A good template plus twenty minutes will get you there.

Use ours when: It stops being sensible at volume. Twenty local businesses, each needing a different adjacency argument and different numbers, is where the template approach quietly turns into never doing it. That, and a template cannot tell you your own proposal is too thin to send.

Who this is built for

Small and medium businesses with no marketing budget

This is the clearest case on the page. Referral partnerships are the cheapest customer acquisition an SMB has available, they need no ad spend, and referred customers arrive already trusting you. Almost nobody does it, and the reason is always the same: working out who to ask and what to say is a day of work that never gets scheduled. Removing that day is the entire point.

See how small businesses use Inflowave and what it costs for a small team.

Agencies, as a service to sell

Setting up local referral partnerships is an easy thing to sell, because the value is obvious to the client and the work is visible. Run the sweep for their area, produce ten proposals, and you have a deliverable that is unmistakably yours rather than a report of numbers they half understand. Proposals carry your branding and stay scoped to that client workspace.

More in the agency use case.

Creators and coaches looking for collaborations

The content collaboration, newsletter exchange and social collaboration shapes are the relevant ones here, and the discovery side works on creators rather than locations. The awkward part of a collaboration pitch is always the numbers, because overstating your audience is both tempting and instantly checkable. A proposal that marks its estimates as estimates reads as more credible than one that does not.

See how creators use Inflowave.

Who should not bother

If you are sending one or two partnership proposals a year to people you already know, a document and half an hour will do the job and this is over-engineering. And if you already have a network of partners and what you need is to pay them commission and track attribution, that is a partner-management platform, which this is not.

How it goes

  1. 1. Sweep your area or your niche

    A radius around a location for a local business, or hashtags and seed accounts for a creator. What comes back is both the competitors and the adjacent businesses, which are two different lists you want for two different reasons.

  2. 2. Connect what you already know before generating

    This is the step that decides whether the output is strong or thin. Real customer numbers, real average order values, anything the client has told you. Every one of those turns an estimate into a verified figure, and the readiness score moves accordingly.

  3. 3. Pick the shape, then generate

    Two-way referral unless there is a reason not to. If the other business is much larger than you, a customer discount or an employee benefit is an easier first yes than asking them to send you leads. Generation runs in the background and takes a couple of minutes.

  4. 4. Read the estimates and correct them

    Every estimate shows its assumptions. Where you know better, edit it. This is five minutes that visibly improves the document and it is the step people skip, then wonder why the proposal reads generically.

  5. 5. Clear the gaps, then send

    Work the readiness list rather than overriding it. If it says thin, it is thin, and the business owner reading it will reach the same conclusion in less time and with less charity than the scorer did.

Common questions

What is a referral partnership?

An agreement between two businesses to send each other customers. Usually the businesses serve the same person at different moments, so the referral is genuinely useful rather than a favour: a bridal shop and a hair and makeup studio, a car dealership and a mortgage broker, a physiotherapist and a gym. It can be two-way, one-way with a fee, or a share of the revenue. The reason it is worth the effort is that a referred customer arrives already trusting you, so they close faster and cheaper than anyone an advert brings.

How does it find partners?

From the same sweep the competitor research runs. It maps the businesses operating around you, then surfaces the adjacent ones, the businesses serving your customer at a different moment rather than competing with you for them, with a stated reason for each pairing. You get real named businesses with websites, ratings and review counts, not a list of categories you then have to go and research. How the research sweep works.

Why are some numbers in my proposal blank?

Because they could not be substantiated, and that is deliberate rather than a bug. Market size in currency and audience-overlap percentages are the two that most often come back empty, since working either out honestly needs data neither business has handed over. Every generator on the market fills those in. Ours refuses, because a made-up number is the thing that loses you the partnership when the other side checks it.

How do I know which numbers I can rely on?

Each one is labelled with where it came from. Figures pulled from your own CRM, given to you by the client, taken from a verified Google Business Profile, drawn from a past partnership, or computed from those, show as verified. Anything from an industry benchmark or an AI estimate is marked as an estimate, carries a confidence level, shows the assumptions behind it, and can be edited by you. The distinction is visible on the document rather than buried, so you know which lines you can defend in the meeting.

What is the readiness score?

A judgement on whether the proposal is actually worth sending, rated strong, fair or thin, with a list of what is missing and how to fix each one. A thin proposal is usually thin because it is mostly estimates: you have not connected the CRM, or you have not told it anything real about the partner. The fix list is specific rather than a general nudge to add more detail.

Can I send one anyway if it says it is thin?

Yes, but you have to type a reason first. It is a speed bump rather than a lock, and it exists because the failure mode is not malicious, it is a person in a hurry firing off something half-built to a business owner they will only get one shot at. Having to write a sentence explaining why is usually enough to make you go and fix it instead.

What do I actually get at the end?

The proposal itself, a shareable web version, the outreach message to send it with, an outline of the agreement, and an email version. All of them derive from the same underlying document, so the numbers in the email cannot drift away from the numbers in the proposal, which is exactly what happens when three separate drafts get written by hand or by a chatbot.

How long does one take to generate?

Long enough that it runs in the background and tells you when it is done, rather than making you sit on a spinner. A couple of minutes is normal. It is doing research rather than filling in a template, which is the reason it is not instant and also the reason the output is worth sending.

Does this replace PartnerStack or an affiliate platform?

No, and it is not trying to. There is no commission tracking, no payouts, no partner portal and no affiliate attribution here. Those platforms run a partner programme once you have partners. This solves the step before that, which is not having any: finding who to approach and producing something worth sending them.

Can an agency run this for its clients?

Yes, and it is one of the easier things to sell as a service, because local referral partnerships are obviously valuable and almost nobody has the time to set them up. Proposals are scoped per client and workspace, and can carry your branding rather than ours. What we build for agencies.

What happens after I send it?

It moves through a status you can see: sent, viewed, then accepted, rejected, or a revision requested. That matters mostly because partnership outreach is slow and easy to lose track of, and a proposal sitting at viewed for three weeks is a follow-up nobody would otherwise remember to make.

The cheapest customers you are not getting

Find the businesses near you worth approaching, and send each one a proposal whose numbers hold up when somebody checks them.

Get Started