Gains:
- Ability to understand basic concepts such as media plan, reach, frequency, CPM/CPC and use artificial intelligence to produce channel and budget scenarios
- Ability to create channel mix draft and distribution scenarios with artificial intelligence support according to target audience, target and budget
- Ability to distinguish that the budget distribution suggested by artificial intelligence is a scenario and needs to be verified with real cost and platform data
Even the best advertising goes to waste if it is shown in the wrong place and at the wrong time. The plan that determines "where, to whom, with what budget, and how often" the campaign will appear is called a media plan. Media planning is one of the most quantitative and strategic areas of advertising because it is directly related to money: it is necessary to allocate a limited budget to the channels that will bring the most value. AI is a powerful scenario generator in this area: it can quickly model different budget distributions, channel mixes, and audience matches. But the recurring principle also applies here: the budget allocation suggested by the AI is an initial scenario, not a definitive plan ready to be implemented. It is not implemented without verification by actual costs and past performance.
Basic media concepts
Before we talk about the media plan, let's clarify a few basic concepts. Reach is the number of unique people who saw the ad at least once. Frequency is how many times the average person sees the ad. Impression is the total number of times the ad appears on the screen. CPM (Cost Per Mille) is the cost per thousand impressions; It is a common measure in branding campaigns. CPC (Cost Per Click) is the cost per click; It is used in traffic focused campaigns. CPA (Cost Per Acquisition) is the cost per acquisition (purchase, registration); It is the measure of conversion-focused campaigns. There is also a goal: does the campaign want awareness, traffic, conversions? The goal determines the channel and metric selection.
The following table summarizes the appropriate channel and metric by goal:
Campaign goal
Priority metric
Available channels (example)
Awareness (new brand)
Access, CPM
YouTube, Instagram, outdoor
Traffic (visit to site)
CPC, click-through rate
Google search, social media
Conversion (sales/signups)
CPA, ROAS
Google search, remarketing
Loyalty (existing customer)
Engagement, open rate
Email, social media
One more concept is important: funnel budget. Instead of focusing the budget only on the sales phase, it is healthier in the long run to distribute it evenly to the awareness (top of the funnel), evaluation (middle) and conversion (bottom) stages. A brand that puts money solely on conversion will struggle to find new customers over time because it cannot nurture the top of the funnel. AI is useful in modeling this balance as a scenario; But the correct ratio is determined by humans according to the maturity and historical data of the brand.
Tip: Clarify the destination before selecting a channel. The right approach is not to put a budget there "just because everyone is on Instagram", but to choose the channel that best suits your goal (awareness or conversion). Likewise, aim to distribute the budget evenly from awareness to conversion, rather than to a single funnel stage.
Step by step: setting up a media plan
Step 1 — Determine the goal and KPI. Awareness or transformation? By what number will success be measured?
Step 2 — Match audience and channel. Which channel does your persona spend time on, and with what intention?
Step 3 — Generate budget scenarios. Model the total budget with different distributions. The AI comes up with a few scenarios here.
Step 4 — Verify with actual cost. Test the scenario with the platforms' actual CPM/CPC ranges and historical campaign data. Cost estimates provided by AI may be out of date.
Step 5 — Allocate and optimize testing budget. Allocate some of the budget to learning; Scroll to the channel indicated by the data.
three mini cases
Case 1 — Speed of script generation. A local restaurant chain did not know how to distribute its 60,000 TL opening budget. They gave AI the goal (awareness in the immediate area) and the budget and asked for three distribution scenarios: one focused on outdoor, one focused on social media, and one balanced. The AI tabulated three scenarios. The team selected the social media + local targeted scenario based on their own regional data and updated it with real costs. AI produced the scenarios, the decision and verification remained with the human.
Case 2 — Price of unverified cost. An e-commerce brand made a plan according to YZ's estimate of "CPM on Instagram is approximately 15 TL" and divided the budget with this assumption. Actual CPM was much higher due to season; The budget ran out earlier than expected and the reach target was not met. Mistake: Putting the AI's overall cost estimate into the plan without validating it with actual platform data.
Case 3 — Value of testing budget. A software brand wanted to put the entire budget into one channel (Google search). With the AI proposal, they allocated 20 percent of the budget to test two alternative channels. Testing showed that LinkedIn returns lower CPA for this B2B product. The next month, the budget was shifted there and the acquisition cost dropped. The testing budget prevented blind engagement.
Four copyable templates
1) Target-channel mapper:
Campaign goal: [awareness/traffic/conversion].Target audience: [persona summary — where spends time].Total budget: [amount]. Duration: [days/week].Task: Recommend 4-5 channels suitable for this target and audience.For each channel: why it is suitable, priority metric (CPM/CPC/CPA), and risk to consider. Giving exact cost; Assume I verify with real platform data.
2) Budget scenario generator:
Total budget: [amount]. Target: [target]. Channels: [list].Task: Generate 3 different budget distribution scenarios (e.g. safe, balanced, aggressive).Make each scenario a table: channel | budget share (%) | Expected role. State at the outset that these are scenarios and will be verified with actual cost and past performance. Don't give an exact number.
3) Testing and learning plan:
Budget: [amount]. Main channel: [channel].Task: Propose a test plan that allocates a portion of the budget (e.g. 15-20%) to learning.(1) 2 alternative channels to test, (2) success criteria for each test, (3) how much time/data will be taken to decide, (4) logic for shifting budget to the winning channel.
4) Media plan summary table:
Convert the following decisions into a media plan summary table:Goal: [...]. Audience: [...]. Channels and budget shares: [...].Task: Table columns: channel, target, budget share, priority metric, measurement method, responsible. Mark the fields I left missing as "to be filled".
Weak prompt / Strong prompt
Weak prompt:
Distribute my 100 thousand TL budget to advertising channels in the best way.
There is no target, audience and duration; AI only gives a distribution with general and fictitious costs, it cannot be verified.
Powerful prompt:
Campaign target: downloads (conversion) of a new mobile game in the first month. Audience: 18-30 years old, users who play mobile games. Total budget: 100,000 TL. Duration: 4 weeks.Task: Generate 3 budget distribution scenarios (table). Let there be channel shares and the role of that channel in each scenario. Let CPA be the primary metric. Exact cost fitting; Assume that each scenario will be validated with real platform data and a 15% testing budget will be allocated.
The second prompt gives the target, audience, budget, and duration; The output is taken as a scenario and verified with real data.
Common mistakes
- Selecting a channel without clarifying the goal. Awareness and conversion require different channels and metrics.
- Mistaking AI's cost estimate for reality. CPM/CPC is current and platform specific; It is not put into the plan without verification.
- Connecting the entire budget to a single channel. Not allocating a testing budget means never seeing a better channel.
- Ignoring frequency. Showing the ad to the same person too often (overfrequency) creates boredom and waste of budget.
- Mistaking the scenario for a plan. The AI output is the beginning; The final plan matures with real data.
Attention: The media plan is a living document. The initial distribution is a hypothesis; Shifting the budget (optimization) according to the direction shown by the data while the campaign is running produces the real value of the work.
In summary
Media plan is the art of distributing a limited budget to the channels that will bring the most value and is the most quantitative area of advertising. Concepts such as access, frequency, CPM, CPC, CPA constitute the language of the plan; The goal (awareness/traffic/conversion) determines the channel and metric selection. AI is a fast scenario generator: it can model different budget distributions in minutes. But the costs and distributions he proposes are a starting scenario; it is not implemented without verification against actual platform costs and past performance. Allocating a testing budget and optimizing based on data is the permanent value of the business.
Application task
Choose a campaign. (1) Write the goal (awareness/traffic/conversion), audience, total budget, and duration. (2) Get 4-5 channel suggestions with "Target-channel mapper". (3) Extract 3 distribution scenarios with the "Budget scenario generator". (4) Select a scenario, verify the proposed costs with actual platform data (by researching) and update if available. (5) Allocate part of the budget to learning with a “test and learn plan”; Note the success criteria and decision time.
checklist
- [ ] I clarified the campaign goal and KPI before channel selection.
- [ ] I matched the audience with the channels they spend time on.
- [ ] I took AI's budget proposal as a "scenario", not a definitive plan.
- [ ] I validated the cost estimates with actual platform data.
- [ ] I budgeted for testing and learning.
- [ ] I included the frequency and optimization plan.