Partnership

There is always a ceiling on what one company can do. A 45-person company ships a handful of significant features a year, and three salespeople meet a few dozen accounts a month. The market doesn't wait for that pace. Customers ask for ten things and three get built. The territory is the whole country and everyone on the team lives in one city.

For a long time there were only two ways to close that gap: hire, or buy. The first is slow and the second is expensive. Five new engineers take half a year to reach full speed, and acquiring a company that already has the technology costs millions. Companies that could do neither simply went without.

That isn't to say cooperation didn't happen. Founders who knew each other passed along leads and ran events together, and they always had. But that kind of cooperation lived inside a personal relationship. It ended when the person moved on, and nothing about what was exchanged, or how much, was ever written down anywhere the company could find it later.

At Pagedeck the ceiling showed up in an enterprise deal. Sofia was working a 500-seat account when the buyer asked whether a video call could be started from inside a document. Theo's estimate for building it was a year: streaming servers, degrading quality gracefully on bad connections, all of it new territory. Grace worked out what that year cost — roughly twenty features that wouldn't ship on the document side. Saying no lost the deal.

Partnership is the third answer to that situation. Instead of building or buying, you borrow. Connect to a company that does nothing but video calls, attach the feature in two months, and you keep both the year and the twenty features. What you borrowed, though, isn't yours. If that company raises its prices or shuts off the connection, one of Pagedeck's features wobbles with it. And the customer who bought the deal on the strength of that feature doesn't care where the failure originated — the support ticket arrives at Pagedeck either way.

So the terms of the borrowing get written down in advance: what it costs, how much notice you get before they change anything, who answers when a customer hits a problem. That is the line between knowing someone in the industry and having a partnership. What to build, what to borrow, and on what terms, is a choice that keeps coming back in the strategy field.

The terms under partnership fall into two branches: how the two companies join up (form), and what the agreement gets written on (documents). Form decides what is exchanged; documents record it.

Form holds four terms, the largest branch here. What separates them is what you hand over and what you get back.

1. Strategic alliance — two companies combining their strengths to grow a market together. It's the broadest of the four, with no fixed thing being traded. Sometimes it's joint development, sometimes opening customer lists to each other, sometimes running events as a pair. No equity changes hands and both sides stay independent, which is what separates it from an acquisition. Being loose, it breaks easily: without a written record of who does what by when, the work stops the moment both account owners change jobs and only the relationship remains. Alliances that hold tend to have a narrow, dated first project rather than a broad statement of intent.
2. Reseller — a partner who sells your product for you. You borrow a sales channel and they take a cut. Rather than staffing a new region, you hand selling there to a company already meeting enterprise buyers in that market. The price is margin. On a contract listed at a million won, a partner taking 25 percent leaves you 750,000, less per deal than selling direct. Companies accept that because hiring into a region and learning it costs more than the difference. The margin also buys something a direct hire wouldn't have on day one, which is a set of buyer relationships that already exist.
3. Integration partner — a partner whose product connects to yours so the two get used together. This is attaching rather than selling, which makes it different in kind from the first two. Opening a video call inside a Pagedeck document belongs here, as does pushing a notification to a chat tool when a document changes. Both sides want their own product used more, so money often doesn't change hands at all. A different cost shows up instead: when the other side changes how the connection works, your feature stops, and customers don't sort out whose fault it is before contacting you.
4. Original equipment manufacturer (OEM) — supplying what you built under someone else's brand. The product carries their name and the customer never learns who made it. You give up the name and get volume in return. The term comes from manufacturing, but software uses the same structure when a large company puts document editing inside its own office suite under its own label. The risks are plain: your brand accumulates nothing in the market, and revenue concentrates in one account. Losing that one contract removes a large share of revenue at once, and there is no direct customer base underneath to fall back on, because none was ever built.

Which of the four to pick depends on what you're short of. Short on routes to customers points to a reseller. Short on capability in the product points to an integration partner. Plenty of both, but a market that won't grow on your own effort, points to a strategic alliance. Product in hand but no time to build recognition points to OEM. Nor does one company mean one form — a relationship that starts as an integration often widens years later into each side selling the other's product to its own customers. That progression is common enough that some companies treat the integration as an audition, watching how the other side handles a small shared obligation before agreeing to a larger one.

Documents holds one term. A memorandum of understanding (MOU) confirms that both sides intend to work together, before the real contract exists. It carries a direction and rough roles, and usually leaves out amounts, deadlines, and named owners. Apart from a confidentiality clause, very little in it is enforceable. It still gets signed for two reasons. Each side needs a document to point at when requesting budget and people internally, and the months before a contract is signed need something holding the discussion in place. The obligations arrive later, in the contract itself, which is where amounts, terms of termination, and the handling of shared customers get settled. Enough agreements stall before that point that many companies write the deadline for exchanging contract drafts into the memorandum itself.

The two branches run in order. The form gets decided, then that content goes into a memorandum and then a contract. Reversing the order causes trouble. Sign first without settling what you'll do together and all you get is a photo of the signing, plus an internal expectation that something is underway when nothing is. Settle the form and skip the paperwork and you'll be arguing about commission rates or who owns the customer a few months later.

"Let's not build this — let's integrate. A year if we do it ourselves, two months if we connect."

A proposal to acquire a capability through partnership rather than engineering.

"A 25 percent reseller margin earns us less per deal than selling direct, but it's cheaper than putting people in that region."

Weighing the commission against the cost of standing up a sales team rather than treating it as pure loss.

"We signed the memorandum last month and the draft contract still hasn't arrived."

A signal that the step where intent becomes obligation has stalled.

Pagedeck's video call question ended up as an integration. Theo compared three candidates, and his criteria weren't features but terms: how often each changed the way the connection worked, how pricing scaled with usage, and who a customer reached when something broke. The one with the best features priced by usage, which meant Pagedeck's unit costs would climb with every new customer. On an enterprise plan sold at a flat seat price, that curve eats the margin on exactly the accounts the feature was meant to win.

With the company they chose, a memorandum came first and the contract six weeks later. Grace insisted on one clause during the memorandum stage: at least 90 days' notice before any change to the connection. A year on, that company did change it, and those 90 days let Pagedeck fix things before a single customer noticed. The clause cost nothing to add at the memorandum stage and would have been awkward to raise a year in.

That same year Sofia built a partnership in the other direction, setting up a channel sales relationship with a firm that already called on regional enterprise buyers and handing it the territories Pagedeck didn't visit. The first quarter produced two deals. The cause, as Sofia read it, wasn't the commission — it was training. The partner's reps didn't know how to sell Pagedeck. She built materials, ran two days of training, and the next quarter closed nine. What she wrote down afterward set the bar for later partners: a commission rate in a contract doesn't make anyone sell, and the days it takes to teach selling belong in the plan from the start.

The most common mistake is reading a signed memorandum as a business that has started. Both founders sign, a photo gets taken, a press release goes out, and it all looks settled. But that document usually names no amount, no deadline, and no owner. Nothing is owed if nothing happens, so neither side has a reason to hurry. Count these as results and six months later you have six memoranda and no revenue from any of them. Whether the document says when a contract will be drafted is closer to a real signal than the signing itself.

Treating partnership as growth that costs nothing fails just as often. No hiring, someone else's sales team, someone else's technology — it reads as free. Two things are missing from that picture. First, the other side needs its own reason to sell your product. A partner's reps lead with what they know well, and yours takes time to learn, so it slides down the list. Second, managing the relationship takes people. Materials to write, questions to answer, customer issues to route between two companies. Past five partners, many companies staff someone full-time on nothing else. What a partnership saves is the large cost of starting, not the monthly handling. So the number worth estimating before starting one isn't the revenue it adds but the work it adds. Measure the hours one partner consumes in a month, sign as many as those hours allow, and you'll do better than signing five and neglecting three. A neglected partner is worse than no partner, because their reps have already told a few customers about a product they can no longer answer questions on.