How to Build a Cross-Channel Media Plan: 7-Step Scorecard for 2026
A cross-channel media plan gives every channel one job, sets its budget and frequency limit, and judges all channels against one shared outcome. Build it in seven steps and score each from 0 to 2. A plan that totals fewer than 8 of 14 points is a list of campaigns, not a plan.
Most advertisers do not have a cross-channel problem. They have a collection of channels that each report their own success. Search says it drove the sale. Social says it drove the sale. The streaming campaign says nothing, because nobody gave it a number to hit. Add the three reports together and you have claimed more conversions than the business recorded.
The fix is not a bigger tool. It is a one-page plan written before the money moves. This guide gives you that plan as a seven-step scorecard you can run against your current media mix this afternoon. It assumes a mid-sized budget, a mix of CTV, social and search, and no data science team.
Why the planning method matters more in 2026
Spend is moving toward channels that are hard to compare. The IAB's 2026 outlook, based on a survey of 205 US buy-side media executives fielded between 24 November 2025 and 16 January 2026, projects CTV ad spend growing 13.8% and social growing 14.6%, while linear TV is projected to fall 1.7%. Budget is leaving the channel everyone could compare and entering the ones that report differently.
Gartner's 2026 CMO Spend Survey, covering 401 marketing leaders in North America, the UK and Europe, found that awareness and conversion together take 62.6% of total media spend. Loyalty and retention received under 15%, a fall of 29% since 2024. Teams are funding the top and bottom of the funnel and starving the middle, which is exactly where a cross-channel plan earns its keep.
For the channel-level numbers behind these shifts, see our 2026 digital ad spend by channel benchmarks. This post covers what to do with them.
The 7-step cross-channel plan scorecard
Score each step 0 (absent), 1 (partly in place) or 2 (written down and followed). The maximum is 14. Screenshot the table below, then work through the steps in order.
| Step | Question to answer | 0 points | 1 point | 2 points |
|---|---|---|---|---|
| 1. One outcome | What single business result judges every channel? | Each channel has its own goal | Shared goal, no number | Shared goal with a target |
| 2. Channel roles | What is each channel's one job? | No roles assigned | Roles overlap | One job per channel |
| 3. Budget split | How was spend divided and why? | Last year's split | Split by gut feel | Split by role, with a reserve |
| 4. Audience map | Who sees which channel, and where do they overlap? | Separate audiences per platform | Shared lists, no overlap check | Shared audiences with exclusions |
| 5. Frequency | What is the combined exposure limit? | No caps | Caps per platform only | Per-channel caps plus a combined ceiling |
| 6. Measurement | How will you know a channel added value? | Platform reports only | One shared analytics view | A holdout, geo test or mix model |
| 7. Review rhythm | When do you move money, and on what evidence? | Ad hoc | Monthly, no rules | Fixed cadence with written triggers |
Read your total as follows. 0 to 7 means you are running campaigns, and fixing steps 1 and 2 will do more than any optimisation. 8 to 11 means you have a plan with gaps, usually in measurement. 12 to 14 means you are ready to test incrementality properly.
Step 1: Pick one outcome that every channel answers to
Write one sentence: what result, in what time frame, at what cost? “Generate 120 qualified demo requests per quarter at or below the current cost per qualified lead” is a plan outcome. “Increase awareness” is a wish.
Pick a business outcome rather than a platform metric. Clicks, views and platform-reported ROAS are inputs. If the outcome cannot be measured outside the ad platforms, such as in your CRM or finance system, you cannot settle disputes between channels later. Score 2 only if the target has a number.
Step 2: Give each channel exactly one job
Channels fail in a plan when two of them are asked to do the same thing. The simplest division uses three roles: create demand, capture demand and convert demand.
| Channel | Typical role | Primary KPI | Judge it on |
|---|---|---|---|
| CTV and streaming video | Create demand | Reach and completed views | Lift in branded search and direct visits |
| Paid social | Create and convert demand | Cost per engaged visit, cost per lead | Leads from audiences CTV already reached |
| Paid search | Capture demand | Cost per qualified lead | Qualified leads, not clicks |
| Programmatic display and retargeting | Convert demand | Cost per return visit or sale | Assisted conversions on warm audiences only |
The roles are a starting assumption, not a law. Social can create demand as well as convert it, and a small search campaign can create demand for a brand nobody has heard of. What matters is that you write the role down so you know what failure looks like. If CTV is a demand-creation channel, judging it on last-click sales guarantees it looks poor.
The reason this division works is visible in the research on how channels interact. A Lifesight report circulated in August 2026 claimed that pairing CTV with paid search lifted search conversion rate by 22.3%, and pairing it with paid social lifted social conversions by 8.5%. Treat those as directional only: the report was produced with CTV sellers including Roku, Vibe and MNTN, and it discloses no sample size or method. The direction fits the role map, though. Reach channels create people who later search.
Step 3: Split the budget by role, then hold a reserve
Start with the capture channel, because it has a ceiling. Search spend is capped by how many people search for you. Fund search to the point where extra budget buys clicks at sharply worse cost, then put the balance into channels that expand the pool of people who will search.
Then divide what is left by role. A workable first pass for a business with little existing brand awareness is to weight toward demand creation, then shift toward capture as branded search volume rises. We do not publish a universal split because it does not exist. The right ratio depends on your sales cycle, average order value and how many people already know your name.
Hold back a reserve of 10% to 15% of the quarterly budget that is unassigned. It pays for the test that Step 6 will recommend and the channel that overperforms in month two. A plan with no reserve cannot respond to evidence. For channel pricing to sanity-check your split, use our CTV advertising benchmarks for 2026 and streaming ad CPM benchmarks by platform.
There is no minimum spend or minimum contract on our side, and programmatic is priced on a CPM basis rather than a retainer, so a first plan can be small and still be structured properly.
Step 4: Map who sees what, and where audiences overlap
Build one audience definition and apply it everywhere. If CTV targets “homeowners in the Greater Toronto Area with a household income above a set band” and social targets a different interest list, you cannot tell whether the channels reinforce each other.
Use first-party lists for three jobs: suppress existing customers from prospecting channels, seed lookalike audiences, and build a retargeting pool that CTV and display can both reach. Our guide to audience targeting covers the options, and the cookieless picture is in our guide to targeting without cookies.
The check for 2 points is simple. Name any audience, and you can say which channels reach it and which exclude it.
Step 5: Set frequency per channel, then a combined ceiling
Every platform caps frequency inside its own walls. None of them can see the others. A household can be inside the cap on CTV, inside the cap on social and still exposed to the same message far more often than any single report shows.
A Conviva analysis summarised by ATDB in April 2026 reported that the average viewer sees the same creative more than 14 times a week across fragmented streaming apps; the summary gives no sample or method, so read it as a warning rather than a benchmark. Vendor guidance from Vibe in June 2026 suggests a household cap of 3 to 5 impressions per week on CTV, again without a cited study. Neither number is gospel, but both point the same way: set a lower cap than you think you need, then raise it only if results justify it.
Write three numbers into the plan: a weekly cap for each reach channel, a combined estimate of how many of your target households see more than one channel, and the trigger that makes you cut frequency. Our CTV frequency capping guide covers the CTV mechanics in detail.
Step 6: Choose how you will prove a channel added value
This is where most plans score 0 or 1. Platform reports are produced by the sellers of the ad, and each reports success on its own terms. They cannot all be right.
Pick one method that sits outside the platforms, in order of effort:
- Branded search and direct traffic lift. Track branded search volume and direct visits weekly after a CTV flight starts. Cheap and rough.
- Geo or audience holdout. Withhold one channel from a matched set of regions or from a randomly selected audience and compare results. Cleaner, and it needs a few weeks of data.
- Marketing mix modelling. Statistical, slower, and the right answer once several channels run for six months or more. See our walkthrough on running media mix modelling without a data science team.
For connected TV specifically, read our step-by-step CTV attribution guide before choosing. Score 2 only if one of these is scheduled with a start date.
Want a second opinion on your channel roles and measurement gap? Our free media plan audit scores your current mix against this scorecard and tells you which step to fix first.
Step 7: Set a review rhythm with written triggers
Decide in advance when money moves. A fixed monthly review with two or three written rules beats a weekly debate about whichever dashboard turned red. Examples of triggers worth writing: “cut a channel's budget by a set share if its cost per qualified lead stays above target for two consecutive reviews” and “release the reserve to the channel with the best incremental result from the latest test.”
Do not judge reach channels on a single week. CTV and social need time to build the audience that search later captures, so give each a minimum test window that you agree before launch. Keep the window in the plan so nobody argues for an early cut when the first fortnight looks quiet.
A worked example: scoring a typical plan
Imagine a Toronto service business spending on search and social, with a small CTV test added in September. Search reports 40 leads, social reports 35, and the CRM shows 55 distinct new leads. That gap is the cost of double-counting, and it shows up in Step 1 and Step 6.
Scoring it honestly: shared outcome stated but no number (1), roles overlapping because social is also asked to capture (1), budget carried over from last quarter (0), separate audiences (0), caps per platform only (1), platform reports only (0), ad hoc review (0). Total: 3 of 14. Nothing here requires more budget. Fixing steps 1, 2 and 6 would move the score above 8 within a month, and it would change which channel gets the next dollar.
Common mistakes the scorecard catches
Adding channels instead of fixing roles. A fourth channel with no job makes the plan harder to read and gives you more reports to reconcile.
Judging reach on last click. It will always look worse than search, because search collects credit for demand created elsewhere.
Treating vendor lift figures as your lift. The 22.3% and 8.5% figures quoted above come from a vendor-sponsored report. Your number could be higher, lower or zero until you test it.
Skipping the reserve. Without unassigned budget, every test competes with something already working, so it never gets run.
Cite this data
Cite this data: North American Media Experts, 'How to Build a Cross-Channel Media Plan: 7-Step Scorecard for 2026' (2026). https://www.namediaexperts.com/blog-posts/how-to-build-cross-channel-media-plan-ctv-search-social-scorecard-2026
Methodology: the NAMEX 2026 Cross-Channel Plan Scorecard is our own framework. External figures come from the IAB 2026 Outlook Study as reported by MediaPost on 28 January 2026 (survey of 205 buy-side executives, 24 November 2025 to 16 January 2026), the Gartner 2026 CMO Spend Survey press release of 8 June 2026 (401 respondents, fielded January to March 2026), the Lifesight State of CTV Measurement 2026 report as covered by PPC Land in August 2026, a Conviva analysis summarised by ATDB in April 2026, and Vibe's June 2026 guidance on CTV waste. Vendor-produced figures are labelled as such.
Frequently asked questions
What is a cross-channel media plan?
A cross-channel media plan is a single document that gives each paid channel a defined role, a budget, a frequency limit and a shared success metric. It differs from a list of campaigns because the channels are planned against each other: CTV builds reach, social and search capture the demand it creates, and one measurement method judges all of them together.
How many channels should a small budget use?
Two or three. Each channel needs enough spend to leave the learning phase and produce a readable result, so a thin budget spread across five channels tells you nothing about any of them. A common starting pairing is one reach channel such as CTV or social plus one capture channel such as paid search. Add a third only once the first two produce stable data.
Does CTV really improve paid search results?
Vendor research says so, but treat the size with caution. A Lifesight report published in August 2026 put the paid search conversion-rate lift from pairing with CTV at 22.3%, yet it was produced with CTV sellers and discloses no sample size or method. The direction is plausible because CTV creates branded demand that search captures. The only way to know your own number is a holdout or geo test.
What frequency cap should I set across channels?
Set the cap per channel and then check the combined exposure. For CTV, vendor guidance such as Vibe's 2026 advice points to roughly 3 to 5 impressions per household per week, though that range cites no study. Because publishers rarely share frequency data with each other, the combined figure is always an estimate, so leave headroom rather than running every cap at its ceiling.
Next step: book an intro call with Ryan to walk through your scorecard, or request a free audit and we will score it for you. For the services behind each channel, see programmatic and CTV, paid social and paid search.