Lead Source ROI: Which Channel Actually Wins You Business

Two advertising channels in the same small business, over the same period:

On the measure most people track, these are the same channel. Sixty dollars apart.

Now the column that was missing:

Channel A returned $8.17 for every dollar. Channel B returned $1.38. Same cost per win, three times the outcome, and no report that stops at cost per win will ever tell you which is which.

The figures here come from one worked sample business — 46 leads across nine sources, 14 won — used throughout the client CRM and pipeline guide.

The Full Table

Nine sources, ranked by revenue:

Across everything: $22,440 of spend, 14 wins, $1,603 average cost per win.

That average is the least useful number in the table. Nothing in this business costs $1,603 a win. It is a blend of a channel at $156 and a channel at $5,100, and every decision available to the owner lives in the difference.

Four Things the Table Says

1. The cheapest channel is also the best converting. Website Form: eight leads, five wins — a 62.5% win rate — at $156 a win. It is the obvious place to put more money, and the obvious question is whether it can absorb any. Channels like this are usually capacity-limited rather than budget-limited, which is why the spend on it is $780 and not $7,800.

2. Google Ads is not failing, it is underpriced against its deals. $1.38 back per dollar is not zero — it is roughly break-even against most gross margins, which means it is buying customers at the price it costs to serve them. That is a bidding and targeting problem or a deal-size problem, not a reason to switch the channel off. Kill it and you lose $14,100 of revenue to save $10,200 of spend.

3. Referral converted three leads out of three. No cost, highest revenue, perfect win rate — and completely non-scalable on demand. The right conclusion is not “stop paying for leads,” it is “start asking,” because this is the only line on the table you can influence without a budget.

4. Three sources produced 13 leads and zero revenue. Google Search, Repeat Client and Directory Listing. Before cutting anything: are those leads lost, or still open? A source whose deals are all still in the pipeline has not failed, it has not finished. That single check is where most channel decisions go wrong.

The Two Ratios, and When Each One Lies

Cost per win = spend ÷ wons. Tells you what a customer costs to acquire. Compare it to your average deal value ($19,500 in this business) and to your gross margin. Ignores deal size entirely, which is why it rated Meta Ads and Google Ads as twins.

Revenue per dollar = revenue ÷ spend. Tells you what the channel is worth. Compare across channels, never against a fixed target. Vulnerable to one large deal: Meta Ads’ $8.17 rests on a single $41,200 win, and with four leads and one won, that is a sample of one. Treat it as a signal to spend more and watch, not as an established rate.

Read together they are hard to fool. Read alone, each has a channel it will mislead you about.

Cost per lead, the third ratio, is the one to demote. It measures the top of the funnel and nothing else, and it moves in the wrong direction under exactly the conditions that hurt you most: a channel that floods you with unqualified enquiries will show a falling cost per lead and a rising cost per win at the same time. If you track one number, track the one with “win” in it.

Making the Data Trustworthy

The table above is easy arithmetic. Getting inputs worth running it on is the actual work, and three rules cover it.

Use a dropdown, never free text. Typed sources become “Referral,” “referral,” “ref” and “word of mouth” — four rows for one channel, and a table that quietly understates your best performer. Ten fixed options is plenty.

Ask at first contact, not at close. “How did you hear about us” is a question with an honest answer on day one and a guess three months later. Attribution invented at close reliably over-credits whatever touched the deal last.

Give a channel a full sales cycle plus a margin before judging it. This business closes in 50.9 days on average, slowest win 69 days. A channel switched on six weeks ago has produced nothing but open deals. Its cost per win is currently infinite and that fact means nothing at all.

What To Do On Monday

1. Put a fixed source list on every open and closed deal you have. An hour, once.
2. Put period spend next to each paid source. Most of it is on one card statement.
3. Build the six columns above. Wins, revenue, spend, cost per win, revenue per dollar — and leads still open, so you never judge a channel on deals that have not resolved.
4. Move money toward the cheap high-converting channel until it stops absorbing it, not toward the one with the biggest revenue number.
5. Re-run it once a quarter. Any faster and you are reading noise; any slower and you have spent a season on a channel returning $1.38.

Related: the weighted pipeline forecast for what those open deals are actually worth, and win/loss analysis for why the leads a channel sends do not convert.

Featured on ReadySheetGo

Client CRM, Sales Pipeline & Lead Tracker — $16.99

Thirteen linked tabs and 7,217 formulas, pre-filled with the nine-source business above.

The Source ROI tab takes the source you picked from a dropdown on the lead row, adds your spend per channel, and returns leads, wins, revenue, cost per win and the best-performing source without you touching a formula — including holding open deals out of the win calculation so a young channel is not judged on deals that have not resolved. The Lists tab holds the fixed source dropdown that keeps the whole thing clean.

Alongside: Pipeline and Forecast with weighted values and rolling 30/60/90-day windows, Follow-Ups building overdue and stalled lists from next-action dates, Win-Loss ranking loss reasons by revenue, Clients with lifetime value and a dormant list, plus Leads, Quotes, Activity, Dashboard and Settings.

Works in Excel and Google Sheets. No macros, no add-ons.

Get the Client CRM, Sales Pipeline & Lead Tracker →