How to Forecast Impressions in DV360 Before You Launch

An impression forecast answers two questions before you spend a rupee. How many impressions will this budget buy? And is there enough inventory that matches my targeting to buy them? DV360 has tools for both, and a one-line formula covers the rest.

What tools does DV360 have for forecasting impressions?

DV360 gives you three forecasting tools, and each answers a different question. Use them together with simple CPM maths.

How do you build a forecast with DV360 Plans?

Plans is the main forecasting tool. You describe the audience and dates, add products with budgets, and DV360 forecasts the result.

1. Go to Inventory, then Plans, and click New plan.
2. Enter a name and location, select Digital as the channel, and set the dates. Add age and gender if you want a narrower audience, and a sub-location such as Karnataka or Bengaluru if needed.
3. Click Next: Add products.
4. Add what you want to forecast: YouTube formats, open auction display or video formats, or publisher and deal inventory.
5. For each product, set the location, flight dates and budget. Use Add another to compare the same format at different budgets.
6. Read the forecast, then download the CSV to share with your team or client.

A few details help. If you start a plan from a campaign, it picks up the campaign's planned dates, inventory source settings, age and gender targeting and viewability targeting as defaults. Brand safety comes from the partner and advertiser settings and cannot be changed inside the plan. You can override the forecast cost with a custom CPM, which is useful when you already have a negotiated rate. India is on Google's list of countries with reach forecasts. Where reach is not supported, a plan shows forecast impressions only. For YouTube on connected TV, plans can also show co-viewing metrics, which count extra people watching together on the big screen.

How do you read the forecast curve?

The curve shows how reach grows as budget grows. Early on, each extra rupee finds new people. Later the curve flattens, and extra money mostly buys more frequency on the same people.

Past the solid curve, DV360 draws a dotted line for additional matching inventory. Google's own guidance is that reaching it is often costly and can waste budget. A practical rule is to set the budget near the point where the curve starts to bend, not at the far end.

Compare two or three budgets side by side. If moving from ₹5 lakh to ₹10 lakh on one format adds very little reach, the second ₹5 lakh will usually work harder in a different format or market.

How does the line item spend forecast work?

The spend forecast tells you whether a line item is likely to spend its budget. It appears in the Intelligence Panel while you create or edit a line item with a future start date, and it updates as you change settings.

DV360 replays historical bid requests that match your targeting and applies your bid, budget and flight dates. It uses history based on the flight length, up to 30 days. The result is one of three labels: high risk of underspend, low risk of underspend, or risk could not be determined. On high risk, the panel suggests creative, bid or targeting changes.

It covers display, video and YouTube audio line items, and YouTube & partners line items of the product and brand consideration type. Three limits matter. The forecast disappears once the line item starts serving. It ignores other line items with overlapping targeting, which may compete for the same users. And Google is clear that it is not a guarantee of traffic.

What does the Inventory Availability report show?

It shows the raw supply that matches your filters, not what you will win. Build it from the Inventory Availability template in Instant reporting. Google moved this report from offline reporting to Instant reporting in September 2026.

The main metrics are:

• Potential impressions: how often devices requested ads in your date range.
• Potential viewable impressions: the share of that supply likely to be viewable.
• Unique cookies with impressions: the unique cookies behind that supply. Users who block third-party cookies are not counted, so it undercounts people.

You can segment by browser, operating system and audience lists, including first-party and third-party lists. There are limits. Combinations under 3,000 unique cookies may show no data. New audience lists can take a few days to appear. You cannot combine audiences with age and gender in one report. And the data is kept for 30 days only.

Never copy potential impressions into a plan as a deliverable. A line item only wins part of that supply, and how much depends on your bid, frequency caps and competition from other buyers.

How do you forecast impressions with a simple formula?

The formula is: impressions = budget ÷ CPM × 1,000. CPM is the cost per 1,000 impressions. For views, divide the budget by the CPV, the cost per view. For a rough reach figure on one channel, divide impressions by the average frequency you expect.

Here is an illustrative example for a ₹12 lakh, four-week Hindi-belt launch across Uttar Pradesh, Bihar and Madhya Pradesh. The CPM ranges are the indicative ranges from our own rate card and guides. GST is extra.

The range is wide on purpose. It shows why the maths alone is not a forecast. Next, tighten each line:

• PG line: the agreed CPM fixes the volume. At ₹300, ₹5 lakh buys about 16.7 lakh impressions. Our Aaj Tak advertising guide shows typical deal ranges and minimums.
• YouTube and open auction lines: add them to a DV360 plan with your real targeting to see forecast impressions and the reach curve.
• Supply check: run an Inventory Availability report on the three states to confirm the open auction line has room to deliver.

For reach, take the bumper line at 17.4 lakh impressions and a planned average frequency of 3. That gives roughly 5.8 lakh people. Reach does not add up across channels, because the same person sees several of them. Our post on reach and frequency formulas explains how to handle overlap, and our CPM calculator guide has more worked examples of the formula.

Why do forecasts and actual delivery differ?

Forecasts use past traffic, and the future rarely matches it exactly. The usual reasons for a gap are:

• Overlapping line items competing for the same users.
• Frequency caps that stop delivery once people have seen enough.
• Brand safety and viewability settings that cut supply after the forecast was run.
• Creative gaps, such as only one banner size or approvals that arrive late.
• Low bids in busy weeks. Diwali, IPL weeks and big e-commerce sales bring heavy competition, and CPMs rise with it.
• Small audience lists that are still building.

Deal type matters too. A Programmatic Guaranteed deal commits a fixed volume at a fixed price, so delivery is predictable. Open auction volume depends on winning bids. Our guide to PG vs PMP vs open auction covers the trade-offs.

What mistakes should you avoid?

• Forecasting before targeting is final. Every layer you add later shrinks the number. Our guide to targeting audiences in DV360 shows which layers cut supply most.
• Planning on the far end of the curve. Extra spend there buys frequency, not reach.
• One CPM for the whole year. Festive months cost more than June.
• Mixing GST in and out. Decide whether your budget includes GST before you divide by a CPM.
• Skipping the line item check. A two-minute look at the spend forecast before launch catches most targeting mistakes. Our step-by-step campaign setup guide shows where it fits.

Forecasting is one step in a full plan. Our guide on creating a programmatic media plan puts it alongside objectives, channel mix and measurement.

Next step

For a first draft, ask for a free media plan: a DV360 specialist estimates impressions, reach and cost from your goal, budget, industry and markets, using our current rate card, within one working day. When you are ready to price specific publishers and OTT apps, our premium inventory page lists deal types and indicative CPMs.