Channels and the mix · Channels

Size a channel

Size a search-ads channel from search volume to 12-month margin in ten minutes, name the input its profit is most sensitive to, and decide.

estimateeasyExercise · 3 min · 26 of 60

On this page

An agency proposes search ads on the terms freelancers type when they need to bill a client. The founder wants a yes or no before the meeting ends, which leaves ten minutes for the arithmetic.

The situation

The invoicing app sells a $50-a-month plan at 80% gross margin, so $40 a month, with 5% monthly churn. The course's 12-month payback target makes $368 of 12-month margin both a customer's value and the most one may cost. The agency's keyword tool estimates 60,000 searches a month for terms such as "invoice software for freelancers". At the planned bid the ads would show on half of them, with a 4% click rate at $2.50 a click. A two-week pilot sent 400 clicks and produced 24 sign-ups, a 6% rate, twice the 3% of the course's paid chain. None of those sign-ups has yet had time to move from the free plan to the paid one, so the share who pay is unknown for search; the agency assumes the chain's 20%. The agency charges a flat $1,000 a month to run the account. Paid search and social already bring the app 96 paying customers a month.

The question

Estimate the channel's monthly profit on 12-month margin, name the input that profit is most sensitive to, and decide whether to fund it.

A strong answer · Rubric · Where answers go wrong

Write your answer first, then open

A strong answer

Chain it to the paying customer. 60,000 searches × 50% shown = 30,000 impressions × 4% = 1,200 clicks: $3,000 of media plus the $1,000 fee, $4,000. Then 1,200 × 6% = 72 sign-ups × 20% = 14.4 customers at $278 each, under the $368 ceiling. Value is 14.4 × $368 = $5,299, so profit is about $1,300 a month.

Now the sensitivity. The profit multiplier is 5,299 ÷ 1,299 ≈ 4.1, so a shortfall of about 25% in any one rate (4,000 ÷ 5,299 = 0.755) erases the profit. The break-evens are 15.1% for sign-up to paying, 4.5% for click to sign-up and 3.0% for the click rate. Cost inputs matter less: the click could rise to $3.58, and with the fee fixed the search volume could fall to about 26,000.

The input to name is sign-up to paying. It is the one rate with no measurement for this channel, and 20% borrowed from the course's chain is a guess about different people. The 6% is measured, though on 24 sign-ups it carries about ±5, so 4.8% to 7.2%, just clear of its break-even.

Size it too: 14 customers a month is 15% on top of today's 96. Searches cap it, so it is a useful channel, not a growth plan.

Fund it for two months at $4,000 a month without raising the bid. Read the paying rate on those 144 sign-ups once they have had time to upgrade: keep the channel at 15% or better and cut it below. If the rate comes in near 30%, the next test is a larger impression share, priced as its own increment.

Rubric

Runs the chain from searches to paying customers, values them at 12-month margin, and includes the fee in cost
Computes the profit multiplier or the break-even shortfall (about 25%) and at least two break-even values
Names sign-up to paying as the input to measure because it is unmeasured for this channel, not only because its effect is large
Notes that search volume caps the channel at roughly 14 customers a month at this share
Decides on a capped test with a stop rule stated as a paying rate

Where answers go wrong

"$4,000 for 72 sign-ups is $56 each, under the chain's $67, so fund it." It stops the chain at the event the platform can see. The decision rests on the 20% nobody has measured, and the answer never mentions it.

"The cost per click is the riskiest input, so negotiate the bid down." The click is the least sensitive input here: it can rise 43% before the profit is gone, while a 25% shortfall in any conversion rate removes it.

"It makes $1,300 a month, so buy every search for $2,600." The volume is capped and the extra share is bought in a costlier auction, as paid media basics prices it. The increment needs its own chain, not a doubling of the average.