The question of “how much to spend on marketing” is one that every business should ask itself. The answer is often “what we spent last year plus a bit more”.
Imagine that instead of simply making tweaks from last year’s budget, you had a blank sheet of paper to design a set of activities and costs with the goal of maximising profit growth over a 3-5 year time horizon – a form of ‘zero-based budgeting’.
When I took on responsibility for the consumer brands at CarTrawler, I had to answer this question so that I could build a strategy and budget at a time where my knowledge of each marketing channel was fairly limited, so I approached it from economic ‘first principles’. A recent conversation with a value creation leader at a mid-market private equity fund prompted me to try to set out my thinking and share it.
In trying to answer the question of “how much to spend on marketing” through an economic lens, what we are really talking about is treating the process of acquiring customers as a demand & supply problem. In this problem, ‘supply’ means the number of customers that we can acquire at any given average Cost Per Acquisition (CPA) and ‘demand’ means the average CPA that a business is willing to pay for a given number of customers to meet its goal for profitable growth.
I’m going to describe each of these components in turn, talking through the theory then coming back to the practical application of this way of thinking.
The ‘supply’ of new customers & problem of diminishing returns
A very common issue in business plans that I see when a business is seeking private equity investment is forecast year-on-year growth in the number of new customers alongside a reduction in CPA. There may be one-off factors that may explain this, but on the whole, if you want to acquire more customers you should expect your CPA to increase. The reason for this is simple – to add more customers you will have to start additional marketing activities which are likely to be less efficient and/or more costly than your existing activities (assuming some level of optimisation has happened over time to lead you to these existing activities).
A good parallel to this is the oil cost curve – which plots different sources of oil against their ‘breakeven price’, i.e. the $ per barrel at which it makes sense to ‘activate’ these supply sources. When oil is >$100 per barrel almost all supply sources become economically viable, whilst when <$50 per barrel a range of sources become too costly to extract for the return you will achieve.
Figure 1 - Oil Cost Curve, Goldman Sachs Research, “Top Projects 2022”, April 19, 2022)
In marketing terms you could think about this as starting with word of mouth as your ‘cheapest’ source of new customers on the left side of the curve, moving through organic search, partnerships, through to paid search, then paid social and perhaps with sports sponsorship as the most ‘expensive’ on the right hand side.
I would recommend thinking about what this supply curve looks like for your business – being sure to factor in the full cost per acquisition, including things such as agency & technology costs as well as personnel costs & sales commissions. It can help you to understand whether you are really maximising the potential of your most effective marketing channels before moving ‘up the supply curve’ to more expensive activities. For example – almost every marketing team is leaving ‘money on the table’ with a limited focus on customer advocacy.
Of course in real life, the confidence you could have in such a curve would rely on the level of marketing attribution you had in place to allocate new customers fairly between the marketing channels that originated them. It would also be hard to account for new marketing activity where you don’t yet have the data to really understand the true cost per acquisition.
It is important to note that we are thinking here in terms of average cost per acquisition. Such an average will almost certainly include some areas of marketing activity which have a very high, uneconomic cost per acquisition that you could focus on to find efficiencies - a great example of the Flaw of Averages in marketing.
For some businesses, this curve will also look unusual, for example it might have a finite end – beyond a certain point you can’t acquire any more customers at any ‘price’ (cost per acquisition), because there aren’t any more in the market; or there might be step functions where beyond a certain volume of customers, CPA increases very significantly.
How much to spend on marketing - the ‘demand/supply equilibrium’ for marketing
I’ve asked many management teams: ‘if you could buy customers off the shelf at Tesco, what would you be prepared to pay for them’ – in other words, the maximum CPA you would pay. This is a difficult question to answer, but we can use our demand-supply thinking here to attempt it.
If you’ve already got an understanding of your ‘customer supply curve’, then there are two factors to consider:
Your expected customer lifetime value over a time horizon that makes sense for your business & how to overlay this on your supply curve to understand the profit maximising average CPA; and
Other constraints for your business, primarily the maximum cash that you are able to temporarily invest in customer acquisition, for example as customers may be loss making for an initial period before they ‘pay back’ the cost of acquisition (this isn’t the same as your marketing budget but rather in finance terms the working capital required to fund marketing activities).
Once you have these inputs, you can work out the cost per acquisition you should target in order to profit maximise over your chosen time horizon. For example, let’s assume that customers generate on average £800 contribution before acquisition costs in year 1, and a total contribution before marketing costs over five years - their lifetime value - of £2,500.
We can apply this lifetime value of £2,500 to each point on the customer supply curve, to work out how much overall profit we would make:
Lifetime Value of Customer Cohort = (LTV per customer – average cost per acquisition) * Number of customers.
When we plot this curve on the chart, we identify that the ‘profit maximising’ cost per acquisition is c.£1,650, which will bring us 7,100 customers. At this level, we will spend £11.6m on marketing and an annual cohort of customers will make £6.2m of profit after marketing spend over the five-year time horizon. Targeting a higher average cost per acquisition will lead to more customers but less profit per customer, and less profit for the cohort overall effectively meaning that each extra customer acquired beyond this point is loss-making.
Now, clearly if average CPA is £1,650 but year 1 contribution is £800, these customers are going to be 'loss making' at first - and not every business would be willing or able to support this. Let’s assume that you (and your CFO!) are willing to temporarily invest up to £2,000,000 cash (working capital) in customer acquisition at any point in time. We can this constraint into a maximum number of customers at each level of cost per acquisition and plot it on our chart:
Number of customers = Cash available/(Year 1 customer lifetime value – average cost per acquisition)
The point at which the 'Customer demand curve' line that we've added crosses your customer supply curve reflects the average cost per acquisition you should target, given the cash constraint. At levels of cost per acquistion below Year 1 contribution of £800, you can see that the line doesn't appear on the chart as you aren't constrained by working capital at these levels in this scenario.
In this *highly illustrative* scenario, you would only be able to afford an average cost per acquisition of £1,270, hence your cash constraint would limit your marketing spend rather than marginal customer profitability becoming negative.
In this simple example we are looking at customer ‘breakeven’ over a full year, in practice you might think about this at a monthly level or an even shorter timeframe – but the takeaway is that often from a ‘demand’ perspective, it is often (but not always) the appetite of a business to invest in temporarily loss-making customers that will set a ceiling on the number of customers you are willing to acquire - in particular in business models with subscriptions or other types of recurring/reoccuring revenue.
In reality, other constraints will also come into play, such as the operational capacity of your business to service any given volume of customers, or perhaps that your input assumptions around unit economics will change because beyond a certain point you would end up acquiring less attractive customers.
How to put the theory into practice
I’ve found that this way of thinking about customer acquisition is a helpful way to evaluate marketing strategy and make decisions, rather than something to be used as a precise ‘model’.
As with much economic theory, the real world doesn’t always behave, input data will always be imperfect and relationships will change over time. But if you have a clear enough starting point of your current marketing effectiveness and customer lifetime value, you can apply this approach in a few areas, for example:
Use the concept of the customer supply curve to identify and prioritise improvements to your customer acquisition activities, for example improvements to your conversion rate, demand efficiency eg PPC quality score, or renegotiating partnership terms;
Calculating the overall impact of improvements to customer lifetime value and the resultant change in your maximum cost per acquisition, for example when considering investments that could improve lifetime value;
Answering my favourite question for marketing leaders of ‘where would you spend your next [£1m]?’ by identifying whether simply having a greater willingness to invest in working capital for customer acquisition could allow you to increase the number of customers you acquire profitably; and
Use this approach to zoom in on a specific channel, for example we’ve recently used this approach with a client to deep-dive into non-brand PPC. This helped us to highlight that their average cost per acquisition was well above the profit-maximising level, and a lower cost per acquisition / higher ROI could significantly increase business profitability.
If you’d like to discuss how you can accelerate customer acquisition in your business, please Contact Us.
All views expressed in this post are the author's own and should not be relied upon for any reason. Clearly.
Comments