Most local businesses can forecast foot traffic with a simple spreadsheet and the transaction or booking data they already hold. A footfall forecast gives you a monthly and weekly view of likely busy and quiet periods, which is enough to plan staff rotas, stock orders and opening hours. This guide walks through the method, including three spreadsheet techniques and where an AI model adds a useful range.
What a footfall forecast is and what it is for
A footfall forecast is a projection of how many people will visit your premises in a future week or month. It is not a precise count. It is a planning number. The point is to predict busy periods before they arrive, not to react after a queue forms.
Use a footfall forecast to:
- Set staff rotas so you have enough cover on busy days and fewer people standing idle on quiet ones
- Order stock before a rush rather than running out or over-ordering for a slow week
- Adjust opening hours around known peaks and troughs
- Decide when to run offers or spend on local ads, because a promotion lands better when footfall is already rising
Foot traffic forecasting works best when it is simple enough to update every month. If the method takes hours, most owners will stop doing it.
Start with the data you already have and the events that move it
You do not need a door counter to begin. Most local businesses already hold a usable proxy for footfall.
Collect one number each week or month from whichever of these you have:
- A door counter, if one is installed
- Till transactions, which usually move in line with visits even if average spend changes
- Bookings or appointments from a diary or booking system
- Calls and direction requests shown in your Google Business Profile performance figures
Record the figure in a spreadsheet with a column for week or month. Roll it up into monthly totals so you can compare periods.
Next to the numbers, keep a notes column for what moved footfall. Bank holidays, school holidays, local events, weather, roadworks and your own promotions all leave a mark. A note such as "road closed for two weeks" or "ran a weekend offer" explains a spike or dip that the raw number cannot. Without notes, you will misread one-off events as a trend.
How to forecast foot traffic with three spreadsheet methods
The numbers below are illustrations using made-up round figures. Use your own history in place of them.
Same period last year adjusted for the recent trend
Take the same month last year and adjust it by how much your recent months have moved.
Suppose last March you recorded 2,000 visits. Over the past three months, your visits have been running at about 1.1 times the same months a year earlier. Multiply 2,000 by 1.1 and your forecast for this March is 2,200 visits. This method is quick and works well for businesses with a clear year-on-year direction.
Moving average
Average the most recent three to six-months and use that as the next month's forecast.
If your last three months were 1,800, 2,000 and 2,200 visits, the average is 2,000. Your next month forecast is 2,000. A moving average smooths out weekly noise and is a good default when you have no strong seasonal pattern.
Seasonal index by month or by day of the week
Work out how each month or day compares with your average, then apply that multiplier to the coming period.
If your average month is 2,000 visits and December has historically run at 1.6 times your average, your December forecast is 3,200 visits. You can do the same by day of the week, for example if Saturdays run at 1.3 times your average daily visits. A seasonal index is the best method for seasonal foot traffic patterns such as Christmas, summer holidays or weekend trade.
| Method | Data you need | Effort | Best for |
|---|---|---|---|
| Same period last year adjusted for trend | 12 or more months of monthly visits or transactions | Low | Stable businesses with a clear year-on-year direction |
| Moving average | 3 to 6 recent months | Low | Smoothing week-to-week noise |
| Seasonal index | 12 or more months of history by month or day | Medium | Strong seasonal patterns such as Christmas or school holidays |
| AI model (Ergora AI Forecasts) | At least six-months of monthly revenue or orders from a supported tool such as Shopify or Stripe | Low after connection | A six-month outlook with an 80% confidence band, not visitor counts |
Turn the forecast into decisions and review it monthly
A forecast only earns its keep when it changes a decision. Use the next month's figure to set the rota, place stock orders and choose opening hours. If the forecast shows a quiet fortnight, shift a small local ads budget into that window. See Ergora's guide to local ads on a small budget for ways to spend a modest amount well. If the forecast shows a busy run, plan the extra cover and stock before the rush, and look at how local businesses use AI to increase foot traffic for ideas that fit your area.
At the end of each month, compare the forecast with what actually happened. Note the gap and ask why. A large miss usually means an event you did not record, a trend that changed, or a seasonal pattern that has shifted. Adjust the method, not the whole system. Foot traffic forecasting is a loop: record, forecast, act, compare, adjust.
Where AI forecasting models help and where they stop
AI models are useful when you have a longer history and want a range rather than a single number. They find seasonal patterns across many months and express uncertainty as a confidence band. They do not know about a road closure, a competitor opening nearby or a promotion you are planning. Those still belong in your notes column and your judgement.
Ergora's AI Forecasts tab uses Google's TimesFM, a time-series model from Google Research, and Ergora runs the 200M model on its own servers. It turns a metric's monthly history into a six-month outlook, month by month, with an 80% confidence band. Every workspace sees sample forecasts for eight business areas, including a local foot traffic sample built from example history and refreshed each week. That card is labelled as a sample and is not a prediction about your business. It shows what a forecast looks like.
Forecasts on your own data appear when you connect a supported tool and the metric has at least six-months of monthly history. Supported tools include Shopify for monthly revenue and orders, Stripe for monthly revenue, Xero for financial figures including cash balance, Google Analytics for organic web traffic, plus Meta Ads, Google Ads, HubSpot and Klaviyo. You can see the full list on Ergora's integrations page. For a local business that takes payments through Stripe or sells through Shopify, the monthly revenue or orders outlook is a practical stand-in for how busy the coming months will be.
Ergora does not read door counters, footfall sensors, booking systems or till systems such as Square, so it does not forecast your visitor numbers today. Square is on Ergora's roadmap. The horizon is fixed at six-months. Ergora has a full write-up of how Google's TimesFM powers its AI Forecasts if you want the detail. Ergora's seat plans start at $99 a month and specialist packs are added from $59 a month each. The only offer is 50% off the first month plus a 30-day money-back guarantee, and there is no free plan or free trial. See Ergora's pricing page.
Start recording this week
A forecast is only as good as the history behind it. If you start a simple weekly or monthly log now, you will have enough data to run all three spreadsheet methods within a few months. The method matters less than the habit. Record the number, note the events, compare the forecast with reality, and adjust. That loop will give you a clearer view of your busy periods than any guess.
Frequently asked questions
How do you forecast foot traffic?
Start by recording a weekly or monthly visit proxy such as till transactions, bookings or calls and direction requests from your Google Business Profile. Roll the figures up by month, note events such as holidays and roadworks, then use a spreadsheet method such as last year adjusted for trend, a moving average or a seasonal index. Compare the forecast with actuals each month and adjust.
What data do I need to predict busy periods?
You need at least a few months of weekly or monthly history for a moving average, and ideally 12 or more months for a seasonal index. Useful proxies include till transactions, bookings, door counter readings and Google Business Profile calls and direction requests. A notes column for local events, weather and your own promotions is just as important as the numbers.
How do I count foot traffic without a door counter?
Use a proxy that moves with visits. Till transactions are usually the easiest, because most shops and cafes already record them. Bookings or appointments work for salons and restaurants. Calls and direction requests from your Google Business Profile performance figures also give a monthly view of local interest.
How far ahead can I forecast footfall?
A spreadsheet method can reasonably look one month ahead, and a seasonal index can extend that across a full season or year. An AI model such as Ergora's AI Forecasts gives a six-month outlook when it has at least six-months of monthly history from a supported tool, but it does not read door counters or till systems, so it forecasts a metric such as revenue or orders rather than visitor numbers.
Can AI predict foot traffic?
AI can predict a metric with enough monthly history, such as revenue or orders, and express the outlook as a range with a confidence band. Ergora's AI Forecasts uses Google's TimesFM and gives a six-month outlook with an 80% confidence band, but it does not read door counters, footfall sensors or booking systems, so it does not forecast visitor numbers today.
What is a seasonal index?
A seasonal index is a multiplier that compares a month or day with your average. If your average month is 2,000 visits and December historically runs at 1.6 times that average, December's index is 1.6 and the forecast is 3,200 visits. You can build the same index by day of the week to plan rotas around your busiest days.