The KPIs you just can’t ignore

A short guide outlining the key operating metrics to look out for as an investor, all within the realms of different business models that start-ups might adopt.

Naturally there will be recurring KPIs between the models, which are by no means exhaustive at an operating level, and apart from SaaS, they don’t even touch on financial metrics.

For those investing at the early-stage, KPIs although somewhat limited by historic performance, will be an indicator of traction, product market fit, the overall ability of management to monitor and action their data to ultimately show whether this idea / team can be successful. For those investing at a growth stage, these KPIs will be a little more fleshed out and will actually demonstrate early signs of success and they’ll be the fuel to feed into your engine of a financial model to predict outcomes, model returns and see if this is a worthwhile investment.

The operating KPIs will be split into the following business models:

Mobile, eCommerce, Marketplaces and SaaS.

Why these models? Simply, they’re common in the venture landscape and all have nuances that are worth exploring. As it’s a little bit of lengthy blog though, the key takeaway is below…

Key takeaway

At different stages of the business life cycle there can be very different outputs for the same KPIs. To be able to identify these core operating KPIs throughout different business models is key for deal execution and picking the right investments.
Those at the early stage (generally Seed and Series A) need to see signs of real traction, product market fit and aggressive growth, early unit economics can be somewhat unproven but thought has to be given to them further down the line. As an investor you need to believe in the idea, and although somewhat blue sky or idealistic by nature, the KPIs need to support this future journey.
Those at the growth stage need proven product market fit, clear monetisation and stable unit economics. As an investor you need to understand your returns profile, and understanding this comes from a combination of scaling, capital efficiency and profitability.
The KPIs below all go someway of proving these points and reaching an investment rationale.
And note, these metrics are not exhaustive and all start-ups with different models in different sectors will have tailored KPI’s and put emphasis on certain ones based on customer behavior. Bench-marking is a relative process.
Finally, care needs to be taken when analysing these metrics, start-ups will usually have limited historical data and assuming their data collection is a 100% robust process is often naive. There is a reason why due diligence is conducted … having said that getting hung up on the exact diligence details vs. the overall view on the idea, company and management, can often cause you to miss out on deals. Investing in early stage and growth is a risky game and finding this balance is vital.
Now carry on reading if you want to see more…

The native apps we have on our smartphones; from social media, to games, to productivity tools.

They may monetise in different ways…

  • An initial purchase (£0.79 on the App store)
  • Freemium (free to use followed by an in-app purchase(s) for a superior product or further content)
  • Virtual goods (in-app purchases focused on virtual money)
  • Completely free for the user (revenue would be generated through in-app adds)

…all the models are comparable with similar operating KPIs.
But note KPI outputs across all Apps will be different depending on monetisation policy and consumer activity patterns e.g. social media is checked regularly and with short time sessions but long life spans, games on the other hand have much longer usage sessions but these sessions will be intense for short periods of time when users are hooked but then drop-off a cliff until re-engaged or never again (the latter is more common).

1) Downloads

  • Simply the no. of downloads. The first step for anyone using the app is for them to download it and the chances are if they do download it, it stays on their phone even if no longer used (unless they’re short on memory, then it gets deleted…).

The Why For Investors
Downloads show if the App is gaining traction in a crowded App store and a reflection of whether any launch strategy or marketing effort is working. This is the first step in a long funnel.

2) Active users

  • No. of people who actually use it, clearly a more relevant metric than downloads. These may be daily (DAUs), weekly or monthly. Take a social media App like Snapchat or Twitter, they are more commonly measuring users on a daily basis (admittedly they have publicly benchmark MAUs) where as a restaurant loyalty app would use monthly as a more relevant metric. Either way the time-frame chosen needs a rationale and will be compared to the peer set.

The Why For Investors
Shows if those who have downloaded it actually use it!
c.25% generally won’t use it once, and a further 25% of those won’t use it more than once, then you’re down to the remaining +55%…Investors should care about engagement as it’s those active users that will flow through any business plan or financial model.

3) User growth

  • The next metric is user growth, this can vary depending on the chosen time-frame, but there needs to be growth and aggressive growth at that. If there is no top line growth, then any growth below this is going to be nigh on impossible at the early stages of the business.

The Why For Investors
Early stage investors want to see growth, they can accept limited monetisation or high CAC’s (below) for the time-being but no top line growth is a no-go! Their returns don’t come from leverage or dividends, it’s growth and growth alone.


  • This one will keep cropping up throughout. Granted it’s a slight cheat as it’s multiple metrics in one…. but the ratio shows the value of your customer (over their lifetime) vs. the cost at which you acquired them. A few things to note on LTV:CAC in general, as well as specific points for Mobile:
Lifetime Value (LTV)
LTV = Contribution per user * (1/Churn) + [referral value]
  • All Mobile businesses seek three key things: Monetisation, Retention, and Virality and LTV is the perfect measure of this.
  • The metric represents the measurable value of the customer over their duration. This is usually measured on a contribution/profit basis but sometimes kept at revenue level, this the Monetisation element.
  • In mobile, Contribution Margin will be high so revenue / GM can be more interchangeable, whereas for eCommerce (see below), using revenue would be misleading due to the lower gross margin (e.g. raw material costs).
  • Note this is Contribution per user, not per order/purchase, therefore incorporating any repeat purchases. In the chase of Mobile, e.g. user in-app spend per month or add sale per user per month, depending on the apps monetisation strategy.
  • Clearly, working out the lifetime (in months) of the customer is key to the metric and is simply = 1/churn rate per month. A low churn means customers like your product. This is the Retention element.
  • Churn can be measured in different ways; unit/customer churn or £ churn. But either way it’s the % that leave over the specific period. Churn can also be on a Gross or Net basis which will be more applicable for SaaS but less so for Mobile.
  • Again, specifically for Mobile Apps, they’ll often include a Referral value, the value they attribute to additional customers you’ll show, recommend and refer the App to, this is Virality. Those Apps with powerful Network Effects will have a hefty and somewhat exponential referral value. As you can imagine, it’s a slightly difficult one to calculate so often ignored.
Customer Acquisition Costs (CAC)
CAC = costs to acquire / no. of customers acquired (in a month)
  • Represents the total expense involved to attract a user within a time period. In simple terms, this cost is usually direct marketing costs (e.g. the cost of an ad campaign on Facebook or Google).
  • However, for more accurate measures these could include additional marketing; staff costs, software, and/or other incremental related costs to the business during that period.
  • Note there is blended CAC i.e. with all customers acquired even those organically, but here I’m referring to paid, those directly coming from the marketing push or campaign. Organic referrals are pretty common with Mobile though due Virality and network effects.
  • In the world of apps, cost per install (CPI) is often used instead of CAC, it is exactly the same metric but Install replaces User.
  • One thing to consider — should the business include retention costs or not? Less relevant for Mobile but much more relevant for eComm, particularly if there is a cost involved in achieving the churn rate to calculate lifetime value, but let’s pick this up further down.
  • So by comparing both of these — what does it show? How/if a customer can payback the cost to acquire them and then any incremental profit.
  • 1x means these are equal and > 1x means the customer contributes more than they cost.
  • Clearly this shouldn’t be < 1x for a long period of time should the business want to continue operating. If it is then work needs to be done to reduce CAC/CPI, or monetise in different ways (e.g. increase installation price and/or consider up-selling to users) to boost the Contribution per user or reduce Churn.
  • As a general rule, investors should look for +3x as a minimum as this demonstrates a healthy business.
  • There are clearly other costs of the business beyond gross contribution and marketing spend - other operating costs, head office, app development etc. So if marketing costs associated with acquisition can’t even be covered then there is no chance for the remaining business costs.

The Why For Investors
LTV:CAC verifies the current go-to market strategy and shows the sustainability of current marketing spends. If this can not be sustained by acquired customers (i.e. a healthy LTV:CAC) then the business will never generate financial value and burn through every last ££ invested, they’ll need to bring more investors on-board and dilute your ownership.

If investing in Mobile, the start-up should exhibit good download numbers, high active users, aggressive growth and a sustainable (or at least route to sustainable) LTV:CAC of +3x.

Other helpful metrics for mobile include App search results, average length of session, session intervals, daily App launches, user sign ups and further granularity on certain Apps screens.

Including pure play online retailers (e.g. ASOS, Thread), direct to consumer operators (e.g. Dollar Shave club, Eve) and consumer subscription services (e.g. HelloFresh, Blue Apron).

1) Site Traffic

  • The no. of hits on the website.
  • It’s a pretty raw metric but if you view the route from website view to sale of a product as a funnel, then the more you feed into the funnel the more sales and revenue there will be at the end, in theory in anyway…

The Why For Investors
This metric is the lifeblood of any eCommerce business and shows customers are looking (and finding) the website organically or through marketing efforts.

Note — there are alternatives to this that measure the initial landing on the website such as unique visits, promotional click-through, but site traffic is the first step to understanding the sales funnel of eCommerce. These will all differ by product offering too e.g. is the website a browsing site where consumers spend time deciding if anything takes their fancy with no real intention to purchase when arriving on the site (e.g. apparel), or do they go on with a purpose to buy a specific item(s) (e.g. Amazon), understanding behavior is key but all will need high traffic.

2) Conversion rate

  • The eComm sales funnel has various layers, but in addition to what goes into the funnel at the top (site traffic), you need to know how this is ultimately converted into a sale - the conversion rate. Along the way there is bounce rate, time spent on site, products viewed, shopping basket abandonment etc, but these are all insignificant vs. true conversion rate.
  • Solid Conversion rates are 3–10%, seems low I know, but imagine all the websites you’ve visited and never bought…

The Why For Investors
Traffic can be high but if conversion is low, this needs to be addressed. It could be quality of the website or the offering is just not what customers are looking for, either way, the business needs to show it understands and can manage the sales process and their funnel.

3) Repeat purchase rate

% of customers who repeat purchases in a specific time period
  • The avg. no. of repeat orders per customers is actually built into the LTV, but repeat purchase rate (e.g. the % of those buying again) is a separate metric
  • It is worth addressing in a little more detail due to it’s importance for eCommerce, we need to know if they repurchase, and then understand how many times in the LTV calculation.
  • Repeat purchase rates are pretty unique to eCommerce though, the other models will focus on retention due to their subscribed offering, recurring revenue, one-off purchases or ad-based monetisation.
  • Repeat purchase is essential in eCommerce because eCommerce businesses cannot usually payback the CAC from a one-off purchase, they require multiple. Plus, customers typically demonstrate low loyalty as they can buy from all retailers due to wider offering, particularly in areas like apparel. So understanding when and how many times a customer comes back organically (without prompting/offers) is critical.
  • Different eCommerce businesses will exhibit different repurchase rates; higher value, one-off transactions will naturally have lower purchase rates (e.g. mattresses with Eve are one purchase every eight years, where as ASOS purchases are probably closer to 3–5 orders (with 3–5 items) every year). Relative bench-marking within the sector is required here.

The Why For Investors
Simply, it shows if customers repurchase and if a healthy rate, demonstrates product market. eCommerce businesses cannot afford to acquire customers for a one-off purchase, they need repetition to achieve profitability.


  • The favourite ratio is back…albeit tweaked slightly for eCommerce
LTV = Average order value * [gross margin] * no. repeat purchases * retention time
  • Similar definition to Mobile, but retention time (e.g. years) is the equivalent of 1/Churn and AOV at Gross margin level * no. of repeat purchase is equivalent to Contribution per user.
  • Almost exactly the same as Mobile by definition, however the inclusion of any additional costs which are very common with eComm should not be forgotten. For example, you sign up for a new meal delivery kit and get the first two free, these need to include in the CAC.
  • One thing to consider that I alluded to earlier are the forgotten retention costs? If no. of repeat purchases are included in LTV, yet any costs to retain these are not (e.g. special offers emailed to existing customers), then surely you’re not comparing apples to apples?
  • Well firstly, these costs are often much lower than CAC, 5x the amount to attract than retain in fact. Secondly, the purpose of LTV:CAC is to not measure this, remember the purpose is to see how much can be spent on acquiring new customers. Retention is a separate measure all together, the assumption in LTV:CAC is that repeat purchases are organic by nature. If you are keen to understand what the retention cost is, then cost per transaction is more suitable, but this relates to both new and returning, CAC is per new customer only.

The Why For Investors
If you can make sense of the LTV:CAC then the reliance on unit economics and the business model is significantly improved.
Again, if the business doesn’t have a strong LTV:CAC then questions need to be raised about how to boost LTV and make CAC more efficient. And if this ratio is beyond repair and shows no sign of improving then the business will never make money.
In eCommerce, after Cost of Goods Sold, which will be taken out via Gross Margin, the next biggest expense is SG&A (i.e. marketing) so understanding how sustainable/effective this spend is key for the growth assumptions and shot at profitability. The only place where SG&A plays an important role is in race to the bottom within on-demand marketplaces (e.g. Uber, Deliveroo) and the mass sales force required for SaaS, but we’ll come onto these.

If investing in eComm, the start-up should exhibit high traffic numbers, a solid conversion rate, loyalty through repeat purchases and a sustainable (or at least route to sustainable) LTV:CAC of +3x.

I’m a big advocate for Marketplaces as a business model so I’ll try not to be too biased here. But the demand and supply aggregation they fulfil, streamlining of services/products and overall efficiency the platforms bring to the market is to be admired. Once they reach a critical mass, powerful network effects takeover as demand and supply follow each other and they can scale rapidly. And from a financial perspective they remove any inventory risk and supply cost unlike eCommerce. However, the beauty of the Marketplace can be saved for another blog post.

The marketplaces include online product marketplaces (e.g. eBay, Etsy and the early Amazon) to service, on-demand marketplaces (e.g. Deliveroo, Bizzby, Airbnb and GoFetch), I appreciate the latter are also Mobile, hence care needs to be taken when assessing.

For marketplaces I decided to cheat a little, below is a link to Andrei Brasoveanu’s post on marketplaces, he puts it brilliantly (probably what you’d expect from an investor at Accel…)

The Why For Investors
Current GMV is somewhat irrelevant given the scaling is still to come, but there does need to be a healthy TAM to reflect future GMV. The business model will rely on this future GMV - if the TAM is restricted then implied GMV will be capped and revenue tiny. LTV:CAC and retention behave similar to eCommerce for product marketplaces and for service-led marketplaces, they exhibit similarities to Mobile through Active Users and LTV:CAC. But overall the Unit Economics need to be robust to achieve profitability, the market opportunity big enough and ensure that the business plan is achievable. Note given a marketplace is naturally two-sided, calculations will be different to those shown above.

To quickly assess marketplaces, you need to be understand how liquid it is, once this is achieved the platform will scale up thanks to the virtuous circle — demand and supply will follow each other, constantly bringing in more and more to each side of the market with every successful transaction, in turn, exponentially increasing the customer offering and ultimately driving profitability if the operating KPIs are on track.

Marketplaces can get very big very quickly but they need to hit a critical mass first which is the difficult part, if not already there understanding how this will come about is the first step, more likely a consideration at the early-stage. Then for the growth investors they need know how the platform can monetise, engage and maintain users on both sides, and understand just how big the market can get — either from taking existing market share, expanding the market or creating it

The most commonly invested business model over recent years is probably SaaS, largely because:

a) It’s super attractive for investors (scales so quickly and generates recurring revenue, has high gross margins, with easily measurable churn and therefore cash burn, investment required and returns can be easily valued).

b) Underlying market trends. The growth of cloud computing means SaaS has become more available and valuable to customers. The broaden offering across different sub-sectors within B2B from CRM to Security to Payroll, has resulted in a superior proposition for customers, plus the growing preference for Opex instead of Capex has proven very attractive.

1) Monthly Recurring Revenue (MRR)

Monthly charge * no. of customers = MRR
  • Simplistic calculation above but any customers on higher fees can also be added into this. It’s also worth noting that the MRR growth is also key, Net New MRR is:
New MRR + Expansion MRR — churned MRR
  • Note this Recurring Revenue could also on an Annual basis (ARR), but the relevance will depend on the contract lengths offered by the company. Also note it is different to Bookings, Deferred Revenue or Run Rate Revenue, helpful summary can be found here. So why has recurring been chosen and why is it so important?

The Why For Investors
For SaaS businesses, investment is made up front — the product will be built and sales force employed to then acquire customers and generate revenue. But the customer doesn’t pay back upfront, instead starts paying back on a monthly basis. Assuming the customer stays for a long time, this will be paid back and provide a steady stream of cashflows. However, to measure this the business needs to know what revenue will still be here tomorrow, so excludes any churned revenue, one-offs etc in that month (i.e. reported revenue).

MRR’s will be relative to the product / sector, and taken alone cannot tell you the whole story. But the MRR should be growing with high Net New MRR and when put in context of the other metrics (below) the higher the MRR, the quicker the payback period (below) and quicker route to profitability.

For an investor, an exit will typically be valued based on MRR so having a clear forecast of this is vital to understand returns.

2) Churn

  • As like all the other models, churn rate is critical. However, given the recurring nature of SaaS, churn is even more important and more practical. In comparison to other models — eCommerce and Marketplace businesses repetition will be over a much longer period and Mobile activity is quite sporadic, SaaS customers should naturally repurchase so any churn is a direct action to avoid re-using the product.
  • If there is high churn (double digits) for the SaaS business, there’s something fundamentally wrong with the product and feedback from customers’ needs to be gathered. Mid to high single figures (5–7%) is generally at the upper limit of acceptable churn.
  • In MRR growth we included the MRR churn in (£) terms, but here are looking at customer churn. There can also be Gross and Net Churn, both of which are focused on revenue churn not the customer and the latter including an up-selling to existing customers, ideally SaaS companies should aim for negative churn whereby up-selling revenue > than churned revenue.

The Why For Investors
Churn shows whether customers like the product for a start. Secondly, at the growth stage, churn is essential for any financial forecasting (below). Thirdly, churn can easily be bench-marked vs. peers, very helpful for investors. Fourthly, given there is so much upfront cost in SaaS, retaining customers is vital in order for them to payback over time. Which leads nicely onto the final metric, once again it’s LTV:CAC…


  • As we’ve seen LTV:CAC shows the unit economics of businesses, none more so than in for SaaS, it is perhaps the most relevant and most applicable user of the ratio. It’s helpful for all models but given the recurring nature of SaaS, importance of understanding customer payback is essential.
  • We’ve already spoken about Churn and MRR, which make up LTV
LTV = average MRR per customer * Gross Margin * (1/monthly churn)
CAC is as previous models:
CAC = acquisition spend / no. of customers won (in a month)
  • For SaaS the majority of costs lie in SG&A i.e the sales force required to acquire the customers, so if these can be covered, then the overall profitability will be close behind.

Again +3.0x is needed here. From LTV and CAC we can also understand payback period on the account e.g. CAC / (Average MRR per customer * Gross Margin).

The Why For Investors
Unit economics are needed to understand if and when a customer will pay back the acquisition costs. If they can’t (LTV:CAC<1.0x) then perhaps avoid investing in the start-up, and if they can then understanding when and how much across their lifetime, will show aggressively the business can be grown.

As Tom Tunguz recently wrote, from these basic metrics investors can determine the cash burn pretty easily.

If investing in SaaS, the start-up should exhibit growing MRR, with a low and improving churn rate and a sustainable LTV:CAC of +3x.