Annual Planning Part 1: Customer Forecasting

How to create customer acquisition and retention forecasts based on your annual revenue goals.

This was originally published in the No Best Practices newsletter on 11.14.2021.

Ok, so it’s (checks calendar) the day before Cyber Monday. Hopefully your holiday season is meeting expectations and the rest of the year should be smooth sailing.

Now is the perfect time to start thinking about planning for next year if you haven’t done so already. But based on my last Twitter poll, a lot of you are done with 2022 planning, which is great! I think you’ll still find the next two newsletters useful. They will add a new dimension to your planning and help hone your strategic focus for the year.

Part one is going to explain how to create customer acquisition and retention forecasts based on your annual revenue goals.

What Is Customer Forecasting And Why Do It?

Let’s say you want to drive $10M in revenue next year. You probably have a merchandise plan tied to that number, as well as a marketing channel forecast. But do you know what percent of those sales will be coming from new customers vs returning customers?

It might feel impossible to predict that, but it’s not! Returning customer behavior is fairly predictable. I go in-depth on the whys and hows of that here, but this is the gist:

• Most of your new customers won’t come back.
• If they do come back, their second and third purchase will be similar to their first.
• By the fourth or fifth purchase, they’re actually more likely to stay than lapse
• A given customer is unlikely to purchase a from given category at a higher AUR in the future

There are exceptions to all of these rules, but those exceptions are few and far between. This means you can estimate your returning customers’ contribution to next year’s sales with a decent level of accuracy.

Why does this matter? Because, as the pundits love to say, it’s cheaper to retain a customer than to acquire a customer. So if you want to hit $10M and your returning customers are only projected to contribute $5M, you want to be sure that acquisition is adequately funded and your marketing strategy is aligned.

Compare these two mandates:
“I need to make $10M with a combination of Facebook ads and emails”.
“I need to acquire 33k new customers at an average cost of $35 per customer”.

Which one is more actionable? You can already see how this process will make your marketing strategy more effective.

Returning Customer Forecast: How To Do It

There are multiple ways to create an accurate forecast of returning customer sales. But I’m going to assume that you have the bare minimum of resources–no coders, no data scientists, no fancy reporting. If your data set is small enough, you can even do this in Excel.

Here are the steps:

1. Gather all of your transaction data for the past 3-5 years*. Tag each order with the year of the customer’s first purchase. You may need to perform some data manipulation to get this info.
2. Sum the spend and unique number of customers by each first order year. You would want to see how much the customers you acquired in 2017 spent in 2017, 2018, 2019, 2020 and 2021 to date. And so on for each acquisition year.
3. Calculate what percentage of customers from each annual cohort came back in each subsequent year, and what their spend per customer was each year.
4. Based on the patterns you observe, project what percent of each annual cohort will come back in 2022 and how much they will spend.
5. You’ll also want to project that a certain percent of the customers you acquire in 2022 will come back again in 2022. You’ll need to make this part of the model feed off your new customer estimate for the year.

Here is a visual of an annual cohort analysis I pulled off the internet to help make it more “real”:

Once you run this analysis you will likely observe the following pattern: for every 100 customers that are acquired in year 1, 25-35 return in year 2 and 15-20 return in year 3. After that point, the decay levels off. The customers that do return spend more in years 2+ than they did in year 3.

*If you have less than 3 years of data, you can analyze what percent of monthly or quarterly customer cohorts have returned.

Evaluating Risk Factors In Customer Forecasting

The forecast you just created is only valid if the conditions in your business remain relatively stable in 2022. So your next step is to get to know your returning customers and determine if any part of your 2022 sales plan or assortment strategy might impact your forecast.

For each annual cohort, pull the following data for each customer:

• What was their first purchase product and/or category, and what was their most recently purchased product and/or category?
• What was their first purchase AUR, and what was their most recently purchased AUR?
• What was their first purchase promo percent, and what was their most recently purchased promo percent?
• Did they make their first purchase during an event or sale? You can track this using promo code redemption or the date of the first purchase.
• What are the most popular category/AUR/promo combinations for all returning customers that purchased in 2021?

Sum up each data pull above–for item 1, your output would be a list of product categories with the total number of customers who made their first purchase from the category.

Then sort them from highest customer contribution to lowest. Calculate each item’s percent contribution to total. Then highlight the first N items that cumulatively contribute to 50% of the total customer count.

Here is a completely made-up example of what one of these charts would look like:

What you’ll typically find is that 10-25% of the items in each list contribute to 50% of the customer volume. These are your risk factors. If there are major changes planned for these price points, categories or events in 2022, you’ll need to forecast your returning customer sales down.

Examples of changes:

• Discontinuing or changing the formulation of a best-selling product.
• Raising the average price point in a category.
• Discontinuing or reducing the offer during a popular promo event.

Planning for Improvements

Of course, there are ways to improve your sales contribution from returning customers. But there is a limited upside to these efforts.

If your average YoY customer retention rate is 28%, you may be able to improve it to 30% with a lot of effort. But you won’t improve it from 28% to 45%. You can get some additional mileage out of increasing the average spend from each returning customer.

Here are some things you can do to improve retention rate and sales per customer from your retained customer base:

• Lower your average AURs or increase the depth and/or number of promotional events.
• Institute marketing programs that convert more of your first customers to a second purchase.
• Improve the relevance and desirability of your assortment, especially in categories that your retained customers buy from frequently.
• Improve the relevance of your marketing to your retained customer base or contact them more frequently.

If you make one of these moves you have to do it strategically because most of these can eat into your profitability.

Customer Acquisition Planning

At this point you have an estimate of how many retained customers will shop in 2022 and how much they will spend. The gap between that number and your total sales goal for the year needs to come from customer acquisition. And you need to drive that acquisition without sacrificing profitability (in most businesses).

To get a baseline acquisition cost, divide your total marketing spend in 2021 to date by the total number of new customers you acquired to date. This high level estimate of 2021 CAC can be used to ballpark the marketing budget you’ll need in 2022. There are a few things you’ll need to consider to refine the estimate:

The Impact Of Merchandise: launching a product that people really want lowers your CAC, because you have to do less work to convince the average person to purchase. This is more of a risk if your assortment changes frequently. But the longer a product has been in your assortment, the more its ability to lower CAC will decay.

Macro Trends In Your Key Acquisition Channels: If you rely heavily on a particular acquisition channel you need to monitor the popularity of the channel. Generally, more advertisers = more competition = higher media prices. You also need to monitor consumers’ relationship with the channel. Catalogs, magazine advertising and linear TV were all “hot” acquisition channels at one time. Eventually Facebook ads will face the same fate.

Wins That Are Hard To Replicate: These are usually PR wins that blow up unexpectedly. Examples: a celebrity wears your product during a highly scandalous moment or your product is featured in a viral Tiktok. Events like this lower your average acquisition cost, but the effect doesn’t last forever and it is almost impossible to reproduce with any predictability.

If you feel like your merchandise outlook is weaker this year vs last, you should increase your estimated CAC. Ditto if there are major headwinds in your key acquisition channel(s) or if you scored a major win in 2021 that will be hard to replicate in 2022.

Your estimated 2022 CAC x Your customer acquisition target = The bare minimum budget you’ll need to hit your financial goals for the year.

Putting Your Customer Forecast Together

As you’re probably aware, growing a business profitably is hard work. The mandate from finance is typically: drive more growth with less resources. You can’t do that by running the 2021 playbook, especially because the macro environment is becoming more competitive every day.

When you complete the forecasting exercise outlined here you are very likely to find that customer acquisition is under-funded for 2022. But that’s a good thing: you now have the foresight to understand what needs to change if you want to be successful.

When confronted with this scenario, most businesses pursue one of three courses of action:

1. Lower growth expectations (rare).
2. Cooperate cross-functionally to address CAC levers across the entire business (rarer).
3. Bury their heads in the sand and expect marketing to “make it work”. (most common)

My next newsletter is going to cover option #2: how to set targets for return on media investment, and how to find more breathing room for your media spend by looking up and down the P&L.