Where Your Deals Actually Come From (And Why You Cannot Tell)

Ask any founder of a $2M to $3M ARR SaaS company where their deals come from and you will get a confident answer. Referrals. Outbound. A partner channel. Content marketing. Founder network. LinkedIn. Whatever the answer is, the founder will say it with confidence.

Now ask the follow-up question. What is the cost per qualified opportunity from that channel, what is the close rate, and what is the average deal size? Watch what happens.

The confident answer becomes a stall. The founder pulls up a dashboard. The dashboard shows leads, not opportunities. Or it shows opportunities but not by source. Or it shows source but the attribution is wrong because the CRM was configured in year one and never updated. Whatever the specifics, the founder cannot answer the question with precision, and the inability to answer it is the reason the company is spending money on channels that are not producing while under-investing in the channels that are.

Channel attribution is the pillar of a revenue system that founder-led SaaS companies neglect most consistently, and the neglect is expensive in ways that do not show up until the company tries to scale.

Why Founders Cannot Tell

The problem is not that founders are lazy about the data. The problem is that at $1M to $3M ARR, the visible signal is misleading enough to feel accurate.

A founder who came out of the network is closing deals from the network. That founder assumes the network is the channel, and in a sense it is, until it is not. Somewhere between $2M and $3M ARR, the network runs out. The founder has not been building the next channel because the current channel has been producing, and the moment the current channel softens, the pipeline collapses in a way the founder did not see coming.

A founder who invested in content marketing is closing deals that came through inbound. That founder assumes the content is the channel, until the founder looks more carefully and realizes that the buyers who closed found the content after they were already looking, meaning the content confirmed a decision the buyer had already started to make rather than producing the decision itself. The channel that actually produced the decision is somewhere upstream, and the founder does not know where.

A founder who ran outbound with a strong SDR team is closing deals from outbound. Then one of the SDRs leaves, or the outbound tool changes its terms of service, or the target buyer becomes unreachable through cold email, and the pipeline drops by 60% in a quarter. The founder learns, painfully, that what looked like a channel was actually one specific person or tool inside the channel, and the underlying channel was never the durable asset the founder believed it was.

In every case, the founder's confidence about the channel was based on the visible output. The output was real. The attribution was wrong.

What Channel Attribution Actually Requires

A working channel discipline has four properties, and the absence of any one of them produces the pattern above.

Every opportunity in the CRM has a source captured at creation, and the source is specific. Not inbound as a category. The specific inbound path. Not outbound as a category. The specific outbound sequence, campaign, or SDR. The source cannot be changed after the deal closes to make the reporting cleaner. The source is captured at the moment of first contact, and it stays.

Every source has a cost associated with it. Content marketing has a cost. SDR outbound has a cost. Paid ads have a cost. Referral programs have a cost. Founder network has an opportunity cost, which is the founder's time. Every channel is measured against the fully loaded cost of the effort that produced it, not against a wishful marketing spend number that ignores the human hours.

Every channel is measured on three metrics, not one. Cost per qualified opportunity. Close rate. Average deal size. A channel that produces cheap leads that never close is not a cheap channel. A channel that produces expensive leads that close at 40% may be the most valuable channel the company has. Founders who only measure the top of the funnel are consistently deceived by channels that look busy and produce nothing.

The channel mix is reviewed at least quarterly and decided on with intent. Which channels are being invested in. Which channels are being tested. Which channels are being phased out. The founder writes the decision down and defends it. The decision is not made by inertia, which is how most founder-led companies make their channel decisions.

Companies that operate against these four properties can defend their pipeline architecture in one paragraph. Companies that do not are running channel investment on gut, and the gut is expensive.

The Cost of Not Knowing

The cost of not knowing where deals come from shows up in three ways, none of which appear as a channel line item on the budget.

The company over-invests in channels that look busy but produce nothing. Content that no buyer reads. Events that produce logos on badges but no logos on contracts. Paid ads with high click-through rates and no attributable revenue. The company spends against these channels because they are visible, and visible things feel like activity, and activity feels like progress. The channels that are actually producing get under-invested because they are not as visible, and the compounding effect of that under-investment shows up 12 months later as flat pipeline.

The company cannot make a case to a serious investor about how it plans to scale revenue. The investor asks the founder how the next $5M of ARR is going to be acquired. The founder says something about doubling down on what is working. The investor asks what is working, in dollars per opportunity by channel, and the founder cannot answer. The investor concludes, correctly, that the company does not have a plan. It has a hope. Hope is not a plan, and hope does not get funded at a premium multiple.

The company misfires when it hires against the wrong channel. The founder hires an SDR team because outbound is working, when what is actually working is one particular SDR with an unusual personal network. The founder hires a content marketer because content seems to be driving inbound, when what is actually driving inbound is a single podcast appearance from 18 months ago that continues to produce warm leads. Every hire made against the wrong channel is a 12-month bet on a false assumption, and the money that pays for the hire is not recoverable when the assumption breaks.

What to Do Monday Morning

Pull the last 20 closed-won deals. For each one, write down two things. The specific source that produced the first meaningful conversation with the buyer. The estimated cost of producing that first conversation, in dollars, hours, or both.

Now sort the 20 deals by source. Count how many came from each source. Divide the source cost by the number of deals to get a rough cost per closed deal by channel. Multiply that by the current pipeline coverage ratio to estimate what it would take to produce the next $1M of ARR from the same channels at the current mix.

The exercise will produce two kinds of surprises. Channels the founder thought were producing turn out to be minor contributors. Channels the founder had not been thinking about consciously turn out to be doing most of the work. Both surprises are diagnostic. Both surprises are the beginning of a real channel decision, made with intent, defended in writing, and measured against a standard the company can enforce.

The channel that will produce the next $5M is almost never the same channel that produced the last $5M. Companies that assume otherwise plateau. Companies that face the attribution question directly and act on it, scale.

If you ran the exercise and the source pattern surprised you, there is a conversation worth having.