From channel choice to channel mix
The choosing channels guide ends with a shortlist of channels that each pass the demand, attention, job and economics tests. The mix is the next decision: how they combine. Two mixes with the same channels can perform very differently depending on the jobs assigned, the budget split and the sequence. The mix is also where the strategy's earlier choices land — the segments to serve, the position to communicate, the objectives to hit — because each of those changes which channels carry the weight.
Step one: cover the journey
Lay the chosen segments against the journey stages and check that each stage, for each priority segment, has at least one channel doing that job.
| Stage | Job | Channels that usually do it |
|---|---|---|
| Awareness | Reach the segment with the position | Paid social, video, display, creators, PR, offline media |
| Consideration | Be present when they research and compare | Organic and paid search (non-brand), content, reviews, email nurture |
| Decision | Remove friction and give a reason to act | Brand search, Shopping, retargeting (RemarketingAdvertising to previous visitors or engagers.), landing pages, messaging |
| Retention | Bring them back and grow value | Email, SMS, app, loyalty, community |
| Advocacy | Turn customers into a source of new ones | Referral programmes, review requests, social |
A gap at any stage is a leak. The two most common gaps are no awareness channel (the plan only harvests existing demand, which runs out) and no retention channel (every sale is bought again). The touchpoints guide covers how to find where your own segments actually are at each stage rather than relying on the generic table.
Step two: set the creation-capture balance
Every mix has a ratio between channels that create demand (awareness and consideration) and channels that capture it (decision). The ratio is the central strategic choice in the mix because the two sides have opposite profiles: capture channels show a low CPACost of each conversion. and a clear attribution trail and are limited by how much demand exists; creation channels show a higher immediate CPA and a murkier trail and are the only way to grow the pool.
The balance depends on the situation. A new brand or a new category needs creation-heavy; a business with unexhausted search demand can lean capture; a mature brand whose capture channels have plateaued needs to shift back toward creation. The brand vs performance comparison covers the same trade-off from the measurement side, and the signs that the balance has tipped too far toward capture: rising cost per acquisition in brand searchSearch queries that contain a company's or product's brand name., falling share of new customers, flat brand search volume.
Step three: assign one job per channel
Write each channel's job in a sentence that names the stage, the segment and the measure of success:
- Paid social — create demand among the household segment; success is reach at a sensible frequency and the lift in brand search and direct trafficVisits recorded as direct because they arrived with no campaign parameters and no referrer, whether typed, bookmarked or from a lost source. that follows.
- Non-brand paid search — capture category demand; success is conversions at or below the target CPA.
- Email — retain and grow customers; success is repeat purchase rate and revenue per subscriber.
A channel with two jobs usually does neither well, because the jobs want different targeting, creative and success metrics. Where a channel must do two (a small team running one social platform for both awareness and retargeting), split it into two campaigns with separate objectives so each can be judged.
Step four: sequence the channels
Channels perform better when they are arranged to hand people to each other than when they run in parallel to the same audience.
- Awareness first, capture ready. Launch the demand-creation channels with the capture channels already live, so the searches and visits the creation channels produce are caught rather than lost to competitors.
- Collect at every step. Every channel should feed an owned audience — email sign-ups, site visitors for retargeting, customer lists — so that later stages are cheaper than earlier ones. The owned, earned and paid guide covers the cycle.
- Retarget by stage, not by visit. People who read a comparison page and people who abandoned a cart are at different stages; one retargeting campaign treats them the same and wastes the difference.
- Avoid self-competition. Two channels bidding for the same people in the same auction, or two campaigns with overlapping audiences, raise each other's prices. The campaign calendar shows these collisions in time; audience exclusions prevent them in the platforms.
A furniture retailer's mix for a new sofa range
Segments: first-time buyers furnishing a new home, and existing customers replacing. Awareness for the first segment: short video and paid social with the range's position (modular, delivered assembled). Consideration: non-brand search on the category terms and a buying-guide page that ranks for them; reviews surfaced on product pages. Decision: brand and product-name search, Shopping ads, retargeting of product-page visitors with delivery proof rather than discounts. Retention and the second segment: an email series to past buyers announcing the range before paid activity starts, so the cheapest audience hears first. The sequence: email a week before launch, search and Shopping live from day one, video and social building through the first month, retargeting from the first week. Each channel has one job and one number.
Step five: weight the mix
Once jobs are assigned, the budget is split between channels by expected return at the margin, with the creation-capture balance from step two as the constraint. The budget allocation guide covers the method; the CPA calculator and reach and frequency calculator estimate what each channel's budget buys.
The channel job sheet
One row per channel, filled in before any budget is attached. If a row cannot be completed, the channel does not yet have a place in the mix.
| Channel | Stage and job | Segment | What it must do | Success measure | Hands people to | Receives people from |
|---|---|---|---|---|---|---|
| Paid social | Awareness: create demand | First-time buyers | Make the position memorable to people not yet searching | Reach at target frequency; lift in brand search and direct visits | Search, retargeting, email sign-up | — |
| Non-brand search | Consideration: capture category demand | All | Answer comparison queries better than alternatives | Conversions at or below target cost | Retargeting, email | Social, content |
| Brand search | Decision: capture people who know us | All | Be present and frictionless on our own name | Conversions; impression share | Checkout | Everything |
| Retention: keep and grow customers | Customers | Bring people back with relevance, not only offers | Repeat rate; revenue per subscriber | Checkout | Every acquisition channel |
Step six: review on evidence
The mix is reviewed quarterly against three questions:
- Is each channel doing its job? Judged by its own success measure, not by a single blended metric that favours capture channels.
- Is the whole working? New customers, blended CAC, brand search trend, share of new versus returning customers — the programme-level numbers that no single channel owns.
- Has anything changed? Audience attention moving between platforms, search demand growing or shrinking, a new competitor in the auction, a new product.
Because channels interact, AttributionAssigning conversion credit to touchpoints. reports alone will mis-credit them — last-click credits capture channels for demand that creation channels generated. The analytics cluster covers the models and their limits; at the strategy level, the practical safeguard is to judge creation channels by reach and by the trend in demand, not by attributed conversions. If the review shows most new customers arriving through one channel or one platform, the guide to reducing dependence on a single channel measures that concentration and plans a way out.
Mixes for three common situations
A new consumer brand with no search demand: creation-heavy (social and video for the first segment), search on the brand name only, email collection from day one, retargeting once there is an audience to retarget. A local service business in a searched category: capture-heavy (organic and paid search, the business profile), reviews as the consideration touchpointAny interaction between a person and a brand, paid, owned or earned, before, during or after a purchase., messaging for decision, email or SMS for repeat work, social as a light awareness layer. A mature online retailer whose search channels have plateaued: hold capture at its floors, move the growth budget into reach channels for the priority segment, invest in lifecycle to raise value per customer, and test one new channel per quarter from the reserve. In each case the mix follows the demand situation, not the channel list.
A channel added to the mix and judged too early, or on another channel's job, is a common reason trials fail; the page on a new marketing channel not working works through that and the other causes.
Common mistakes
- A list instead of a mix. Channels with no stated jobs, judged by one blended number.
- All capture, no creation. Efficient until the demand runs out, which it does.
- One channel, two jobs. Awareness and conversion from the same campaign, with the metrics of neither.
- Parallel instead of sequenced. Everything launched at once to the same people.
- Reviewing by habit. The mix stays the same because it is the mix, not because the numbers say so.
- Judging creation channels on last-click. Then cutting them, then wondering why capture channels got more expensive.
For the planning cycle in which the mix is set, see annual and quarterly planning; for the author's approach to planning and buying across channels, see media planning and buying.
