Channel Mix

With one or two channels nobody needed the phrase. You ran an ad, you looked at the result, that was the job. The problem appeared once channels multiplied to six or seven. Each owner optimized for the best number their own channel could produce, and those good numbers did not add up to anything.

Pagedeck spent a quarter living inside that failure. The ad team picked the copy with the best click rate, which was "Start for free." The blog published on the topic that drove the most traffic, which was a tour of the free plan. The lifecycle emails going to signed-up users explained the free plan's features. All three channels reported hitting their goals, and people already using the free tier received the same message three times. Meanwhile anyone weighing whether to upgrade got an answer from nobody.

Measuring each channel on its own hides that overlap. Add up clicks across three channels and nothing in the table reveals that the same person appears three times. A channel mix is the plan drawn up to remove the overlap by dividing roles before dividing money. It moves the question from running one channel well to running several as a single system.

Roles get divided by where a customer already stands. The sequence a person moves through — hearing about a product, comparing options, buying — is the marketing funnel, and channels get placed along it.

1. Write down the question each stage has to answer. Someone hearing about you for the first time is asking what problem this thing solves. Someone comparing is asking how it differs from the alternatives. Someone close to a decision is asking what it costs for a team their size.
2. Attach the channel best suited to each question. Paid media, where you buy placement on someone else's property, can reach people who don't know you yet, which suits the first stage. Owned media — the company's own blog, docs, help center — has room to explain at length, which suits comparison. Email to people who already signed up takes the last stage.
3. Give each channel exactly one number to answer for. Ask a top-of-funnel channel about closed sales and it will start chasing only buyers, abandoning the role it was given. First stage counts newly reached people, the middle counts the share who read a comparison page, the last counts paid conversion.
4. Put budget on top of roles, in that order. Reverse it and money flows to whichever channel produces the cheapest clicks, which is almost always a channel re-reaching people who already know you.

The table only works if one person owns the whole of it. When each channel owner fills in their own row, the roles drift back toward whatever that channel is already good at, and the overlap creeps in again by the second month. Someone has to be able to tell a channel owner that a strong number came from work the plan assigned elsewhere.

Set a review interval at the same time. How credit for a sale gets assigned across touchpoints is called attribution, and a model that hands all credit to the last click makes every early-stage channel look worthless. Move budget on the strength of that table and the channels doing first contact get cut first; two quarters later the whole pipeline runs dry.

Annual budgeting is the first occasion. Fill in the role table before writing any amounts, and where a stage sits empty, start there. A mix with nothing assigned to comparison brings in plenty of people who then stall at the decision.

A new product or a new target customer calls for another pass. A combination that reached individual users does not reach the manager who approves purchases. Here it's often faster to give an existing channel a different role than to swap channels. Keep the blog and shift its subject from personal productivity to team rollout stories.

Budget cuts are where the mix gets used most. Trimming every line by the same percentage weakens every stage slightly and leaves nobody able to say where the break happened. Working from the mix turns the cut into a choice about which role to give up, and teams usually shrink first-stage work and keep whatever converts people already in the door. It's worth writing down that this decision reduces next year's arrivals.

"There's nothing in the mix covering the comparison stage."

No channel is answering the people actively evaluating the product.

"This quarter's mix is 6 paid, 3 owned, 1 email."

Setting the weight of budget and role in numbers.

"That's a top-of-funnel channel, so don't judge it on sales."

Stop measuring a channel with a number that belongs to a different role.

Every quarter Ravi redraws a single-page table at Pagedeck. Stages down the left, the channel assigned to each in the middle, the number it answers for and its budget on the right. Grace reads that table by looking for blanks rather than amounts — whether a stage is missing matters before how many channels there are.

Keeping it to one page is part of the rule. Attach a detailed plan per channel and nobody looks at the whole thing any more, at which point the table built to catch overlap has turned back into a stack of channel reports.

Sofia reads it too. Knowing which question people arrive with tells her which stage isn't holding up its end. A sudden rise in pricing questions on calls reads as a comparison-stage channel that stopped explaining price.

The most common mistake is reading the mix as a budget split. Meetings discuss it in percentages, so the percentages start to look like the whole thing. But a split alone leaves channels doing identical work with different amounts of money — three channels all promoting the free plan, divided 6 to 3 to 1. Once that's the state, moving budget around produces roughly the same outcome no matter where it goes, and the team never finds out which channel is the weak one.

Treating more channels as a better mix is the other one. Being present everywhere feels like missing fewer opportunities. Running a channel properly means producing material suited to that specific place, though, and that work scales with the channel count. Three people covering eight channels end up maintaining all of them at the minimum, and customer acquisition cost (CAC) — what it takes to bring in one customer — climbs separately in each. Teams that drop three and concentrate on five frequently come out ahead.