November 2026 Meta–Google Media Plan: Budget, Forecasts, Guardrails, Phasing, and Reallocation Strategy

1. Give each allocation a job

A useful media plan is more than a Meta–Google split. Each line needs a commercial purpose. Branded search captures existing demand; non-branded search and product-led Google campaigns seek shoppers expressing category interest; Meta prospecting introduces products to potential new shoppers; retargeting addresses people who have already engaged. These roles can overlap in attribution, so they should not be presented as seven independent sources of incremental sales.

Watch how the following walkthrough separates targeting, schedule, budget and forecast in a media-plan sheet. Its example forecasts leads; your deliverable will forecast store sessions, orders and revenue instead.

For a travel-goods store, organize the proposed spend around hard luggage, backpacks and travel accessories, but do not assume all three have equal stock, margin or conversion rate. In the absence of product-level evidence, put the main prospecting effort behind in-stock luggage and backpacks with credible product pages. Use accessories as a cross-sell or a bounded test; expand their budget only if their economics support it.

2. Build the November allocation and forecast

Add a new tab to your Sheet. Use the following illustrative assumptions, replacing them wherever the assignment supplies reliable comparable data. The Google product-led line is a planning bucket: choose the suitable Shopping or Performance Max approach after checking the feed and account. Do not count the same spend in both.

This allocates ₹19.5 lakh to Google and ₹10.5 lakh to Meta, including a clearly identified ₹3 lakh testing reserve within—not on top of—the ₹30 lakh ceiling. The ₹3 lakh branded-search line is a cap, not a reason to buy searches that do not exist. Retargeting likewise cannot be scaled beyond its eligible audience simply because a spreadsheet assigns it money.

Put platform in column A, planning line and role in B–C, spend in D, CPC in E, click-to-session rate in F, session CVR in G, and assumed AOV in H. For this first pass, enter ₹4,500 in H2:H8, matching the previous lesson’s baseline assumption. Use columns I:M for sessions, orders, revenue, cost per order, and revenue divided by spend. Enter these formulas in row 2 and copy them through row 8:

I2: =IFERROR(D2/E2*F2,"")
J2: =I2*G2
K2: =J2*H2
L2: =IFERROR(D2/J2,"")
M2: =IFERROR(K2/D2,"")

In row 9, sum columns D, I, J and K; calculate L9 as =D9/J9 and M9 as =K9/D9. Retain fractional expected orders in the formulas, even if you display whole orders. The illustrative totals should be approximately 94,933 sessions, 1,919 expected orders, ₹86.34 lakh revenue, ₹1,564 media cost per order, and 2.88 revenue-to-spend.

These totals are deliberately close to—but not identical with—the previous baseline of 96,000 sessions, 1,920 orders and ₹86.40 lakh. Row-specific CPCs and CVRs account for the small difference. If your edited plan diverges substantially, explain which assumptions changed rather than forcing the row totals to equal the earlier forecast. Do not add the improved-checkout uplift to these rows unless that change is implemented and you explicitly model its effect across the relevant traffic.

3. Put financial and learning guardrails beside the forecast

Revenue divided by media spend is useful for steering, but it is not profit. You need contribution margin—revenue remaining after applicable product, fulfilment, payment, discount and expected-return costs, but before advertising—to estimate an ad-spend break-even point.

If that margin were 50% purely as an example, break-even would be $2.0$, equivalent at a ₹4,500 AOV to a ₹2,250 maximum break-even cost per order. You might set a more cautious scaling gate of 2.3 revenue-to-spend and no more than about ₹1,957 per order. Neither threshold is a company fact: replace both when the relevant margins, returns and product mix are known. A lower actual margin raises the required revenue-to-spend.

Make the rules operational:

• Maintain or cautiously scale a proven line when at least seven days of reasonably mature data and roughly 30 or more orders support performance above the scaling gate, inventory is available, and the combined store-level trend is not deteriorating. Increase budgets gradually rather than moving the whole month at once.
• Keep the test reserve bounded. Its initial forecasts are below the illustrative $2.0$ break-even gate. That can be justified as a capped learning investment, not described as profitable scale. Define the hypothesis before spending; after sufficient data, move unspent test money to a qualifying line or hold it back.
• Do not force branded search or retargeting to absorb spare budget. Their attractive forecast ratios may reflect existing intent, attribution overlap or a limited audience. Look first for capacity in eligible non-branded, product-led or prospecting activity.
• Treat missing or immature results as uncertainty, not failure. Check order and revenue reporting on a consistent date range; use platform figures to diagnose campaigns, but assess the overall paid plan against the same store-side definitions used in your forecast. Platform-attributed orders may overlap.

The previous downside scenario is the reason for a stop rule. If CPC rises, fewer clicks become sessions, and CVR weakens, the plan’s baseline efficiency may no longer hold. If the combined, sufficiently mature revenue-to-spend falls below the margin-based break-even gate, pause discretionary increases, check tracking and site conversion, and revise the forecast before committing the remaining reserve.

4. Phase November spend, then specify the reallocation decision

November has 30 days. A weekly table makes the budget actionable without pretending that every day will perform equally. Use this conditional schedule; the later weighting is appropriate only if demand, promotions, stock and site readiness support it.

The remaining spend in each week is for the five non-test lines, initially in proportion to their monthly allocations. Seventy percent of the test reserve is scheduled in the first two weeks, leaving time to learn and reallocate. If no late-month demand case is substantiated, replace this schedule with a roughly day-weighted one; do not claim a seasonal uplift without evidence.

At each weekly checkpoint, compare actual cumulative spend with the table’s cumulative target, not with a flat monthly average. Record orders, revenue, cost per order and revenue-to-spend using a consistent reporting cut-off. Then apply this assignment-ready rule:

After allowing for reporting delay and checking stock, move up to 10–15% of a lagging line’s unspent budget per weekly review toward a line that clears the margin-based scaling gate and has room to grow. If no line qualifies, retain the unspent amount and report the likely underspend rather than buying inefficient volume to reach ₹30 lakh.

For example, if ₹15 lakh has actually been spent by 14 November, versus a ₹13.5 lakh plan, only ₹15 lakh remains for the final 16 days—an average of ₹93,750 per day. You would reduce the remaining schedule or explicitly approve a revised cap, not keep spending to the original weekly targets. Pacing is a financial control; it does not override the performance guardrails.

Google’s retail guidance is useful when you turn November evidence into the following months’ plan. Read only the passages on forecasting demand and separating products when different priorities warrant it.

Finish the submission with directional, not fabricated numerical, forecasts for December and January. For December 2026, use November’s mature results, demand signals, stock and margin to decide whether to retain the November mix or shift budget toward a proven product group; keep any promotion-dependent expansion conditional on economics. For January 2027, reset the forecast using actual CPC, sessions, CVR, AOV and returns. Refresh prospecting creative and category priorities, reduce any allocation whose demand has faded, and size retargeting to the audience that actually exists. If the ₹30 lakh monthly ceiling continues, say so as an assumption; do not imply the later months have approved budgets.

What your submitted plan should establish

You now have an allocation that totals ₹30 lakh, assigns Meta and Google distinct roles, reserves ₹3 lakh for defined tests, and rolls row-level assumptions into sessions, orders, revenue and cost per order. Its weekly targets control when money is spent; its margin-based gates and reallocation rule control whether remaining money deserves to be spent. Keep the assumptions visible—especially AOV, margin, product availability and conversion rates—and the plan remains useful even where historical data is incomplete.