SaaS / Cloud KPIs

What are you measuring?

KPIs is SaaS is a good way to measure your business. It is not the only way to measure and you often end up measuring the business looking backward. This is wrong. You NEED to measure the business looking forward. So your backward analysis is to define the trends and you project forward looking KPIs.

Don’t start measuring when you get your first customer. Call them instead and understand. A minimum expectation is to get 100 customers before measuring your SaaS business. Anything else, is measuring the mess and have no trends or validity.

In addition, you create levels of acceptable ranges. For example, churn can be between 1-4% monthly. Anything higher is unsustainable and as you measure you can call the red flags.

Each KPI can be viewed in a range of dimensions/segments.

  • Dates: Often monthly, quarterly and yearly.
  • Channels: License models, sales channels, geography etc.
  • Customer segmentation: size, type, profile etc.

SaaS KPI list

KPI Description Calculation
ACV Annual Contract Value.
  •  Total contract value / 12
ARPU Average revenue per user/unit. An indicator of revenue based on the user or unit base of your business.
  • Total revenue / total number of user/units assigned to the revenue
ARPA Average revenue per account. Unlike ARPU, ARPA is calculated on the number of customers.
  • Total revenue / total number of accounts assigned to the revenue.
MRR Monthly recurrent revenue. RR on a monthly scheme.
ARR Annual recurrent revenue. RR on an annual scheme.
RR Recurrent revenue. Often mistaken by run-rate. Not the same.


Recurrent revenue is the portion of your business, which will auto-bill and reoccur automatically.

CAC Customer acquisition cost. To cost assigned to acquiring a new customer including sales and marketing cost.
CRR Customer renewal rate.
Churn Churn is an indication of a loss in your business. We define churn into different elements:


·         Customer churn

·         Revenue churn

·         Seat/user churn


100 customers at the start of the period, 7 of them cancel by the end. Your churn rate is 7/100=0.07 = 7%


You can use customer churn rate to calculate average customer lifetime if you divide 1 by your churn rate.

By dividing 1/0.07, we see that a 7% churn rate equates to a customer lifetime of 14.2 months.


Revenue churn may not be loss of customer. But simply customers moving to another tier or plan.

Customer churn:

  • Lost customers / total customers


Revenue churn (monthly):

  • Lost MRR / total MRR


Seat churn:

  • Lost users / total users
COGS Cost of goods sold
CLTV Customer lifetime value, also indicated LTV.


The value a given customer delivers to your business over the time.

CMRR Committed Monthly Recurring Rate. More or less a modified version of MRR, the goal of tracking committed monthly recurring revenue is to show what a SaaS company’s revenue stream will be going forward if the business halted its sales and marketing efforts.
LTV Customer lifetime value. See CLTV.




$100 ARPA / 0.1 churn rate = $1000 LTV

Your LTV / CAC should be >= 3x.

ARPA / Customer churn rate:

  • ( Sum of all customer MRR / Total # of customers ) / ( # of customers who churned / Total # of customers )


  • MRR * Gross Profit % * #lifetime in months
PBP Payback Period. The period before any customer is paid of the investment to acquire.

Your PBP should be <= 12 months.

PBP = CAC / MRR * Gross profit %
ROI Return of investment.

This list will be regular updated.


Tools for your customer experience

In an earlier article, I described a model for how to optimize your trial and customer experience. In this article allow me to focus on some real tools for you to drive your business.

  • A customer platform. A no. Not a CRM platform. A customer platform. What’s the difference? Well, if you think CRM – 99.0875% of all of you think sales force and marketing. In SaaS – the service is everything. From sales, to support to marketing, to help, to customer engagement to conference calls and online webinars. Still there is no single handed SaaS platform on the market, so your platform for customer will be multiple apps integrated… CRM, ERP, HRM etc. Choose wisely and consider the cost of integration.
  • Drive your customer onboarding automatically. WalkMe is a set of amazing snippets you design and insert into the framework of your application. WalkMe is a great way to automate “how to use” and “show me” features directly with your SaaS application. The WalkMe use a clever engine to basically do the steps for the user, by integrating into the HTML. I really encourage you to learn the structure –
  • Automate your customer triggers. Marketo is the defacto standard of optimized process management for onboarding and managing both customer sign up and customer journeys. As you open the doors to your SaaS application, you need to design every aspect of customer experience. From signing up, to activating, to getting people on the platform and to billing. Marketo can design these workflow based on the platform, instead of bundling 100’s of apps to drive the same.
  • What are my users doing? I have participated in more than 1.000 reviews and board meetings on the discussion of investments and future strategy for these companies. I always see the same conversation; R&D says “we would like to invest into feature A to optimize our code”. Sales says “well new users need feature C so please help otherwise we cannot sell more”. Marketing says “customers search for B”. And I always ask – “what does the customers tell you?”. The room go silent. All think of the burden if any is tasked to call 50-200 customers and ask. Sales and marketing leave the room as they need to do to “the toilet”. Reality is the CEO/MD failed a complete different place, but lets leave that for now. Telemetry is the key to your insight and above all you need this so emmence.  SaasOptics or provide a platform to engage. UPDATE- Microsoft Azure now provide a telemetry as a service. This is technical and not business – so you need to craft a middleware.
  • Billing and dunning. This is the most important investment you will drive. Probably a key part of your customer platform, but often a separate ERP integrated with multiple credit card vendors. I urge you to use an automated platform example

But wait? You did not explain where the CEO gone wrong in “what are my users doing”. Well please revisit this article, and this. In addition learn the following; in SaaS/Cloud – your product is a service. Your service is your organisation.

There are no siloed organisations where R&D handover to product marketing and they handover to marketing whom handover to Sales. STOP DOING THIS. Your customer is really the center – and there is no boundary. R&D CAN easily own all marketing communication.

UPDATE – Very good article here to review the telemetry designs.

Effective trial experience

In cloud – trials has become the defacto way to attracting interest from your prospects. As in the former shareware days, where TuCows enabled customers to download WinZip, WinAmp and similar applications for a 30 days grace period – trials are a decent want to show case your product.

However, as you design your trial experience – stop for a second and consider the customer experience. In my 10 years of cloud – excessively I have seen one stereo type of trial experience.

  • Dozens of fields for signing up.
  • Captcha to “remove robots” (actually just to annoy you…)
  • 30 days “test period”
  • Regular scripted emails, sent by a clock.
  • Inappropriate follow up
  • Very steep entry to onboarding.

So let me explain. This is wrong. This is completely wrong. Why?

Because your product is not a stereo type product. Why? Because you told me. When I meet you as an investor, as an angel, as an advisor – you tell me how special your product is. How profound and differentiated it is. Yet – when you want to present it to the best asset (not being the moneyman) but the customer – you fall back to the dumbest idea of all.

Copying from the market leaders….

Take a step back

You need to take a step back. Validate your product and the customer experience. Those two words are more important in cloud that you every think. They are the divide between you having 2.000 free comments on a social community – “that you just experienced the best trial every, and bought the product 3 minutes later” – or you desperately call customer support – only to be put on queue.

Cost is a big thing in trial. COGS increase as you drive up capacity in the cloud. The cost shifts towards the vendor (you – my dear) so be sure to report the COGS of trials as a marketing cost. Consider the impact.

100 trials, with a compounded daily cost of $0.3 dollar is nothing. Over a month, it is easy $900 – and 10.000 is $90.000. What happens if customers open duplicated trials? What if 70% call customer support – and the $0.3 becomes $3.0? I have seen this very often – and CSVs (cloud service vendors) often explain the impact as – “cost of success”. I will let you in on a secret. It is not success.

Start from the beginning

Step 1. Leaping into your customer journey starts with R&D. Not rest and drinks – research and development. Tweak your product to be as simple as possible. As connected as possible. Without 40 different tools, portals and applications your need to master. If you can get one interface – you are top of the rock. Spotify did it by creating the sign-up experience within the application. No forms on a website. No captcha. No e-mail… 3 seconds and you listen to Bach.

Step 2. Be honest. Asses the complexity of your product. It is perfectly fair to offer trials of the next Mars Planetarium simulator – if customers would buy it. However, you need to be fair to your customers – maybe ask them how easy the application is to use.

Step 3. Customer data is king! Actually, customer data and users are king! Once you get a customer to upload data, create users – you are well underway in driving the adoption of your service. However, with data comes configuration – and adoption to that data. This often entails customizations (typically for CRM, ERP and HRM) – so understand the timeframe, cost and resources your potential customer need to use to customize the solution. The trial is likely to be used for a POC (proof of concept).

Consider the successful range in days for your customers. Trials below 5 days are good for consumers, but businesses need more.

Step 4. Connect the dots. Why offer 30 days for a very simple solution with no complexity and no customization? Any customer would asset the need of the application within 5-7 days. On the other hand, – the Mars application might need 3 months to monitor the star patterns – and without this, no value, and not value becomes no application. I have seen too often vendors lock the customers purchase timeframe to a fixed number of days. Offering trial extensions is a good approach – however nobody is making it easy due to the COGS. Why?

Step 5. Communicate by the 3C’s. Context, content and channel. Standardized emails are good. However, if you are testing an ERP application – is it financial management? Property management? Assets management? You need to tailor the experience to the expectations of your customer. If you offer a very complex product – why not add a Skype video conference call as the first page upon signing up for a trial? It will enable direct communication to your customer. Use an improved channel.

  • 7 of 10 trial emails are deleted.
  • 1 of 10 trial emails are tagged as spam.
  • 5 of 10 trial emails are revisited to login details.
  • 3 of 10 customers do not want to do it themselves.

What offer do you provide? What do your customer expect?

Use the following Boston 2-by-2 and consider your trial experience


Complexity of the solution


Drive hands-on assisted trial experience

Use dedicated resources

45-90 days, extendable

Use customer success managers

Offer hands-on assisted trial

You need to implement a nurture platform to understand who you would offer it to

30-45 days, extendable

Offer standardized trials

Use an automated approach

15-30 days

Offer self-service trials

Use an automated approach

5-15 days

Low < Volume of expected trials > high

As you design your trial experience – think about how you get the customer in and on. Do they only need to download a client, open a web application or do they need to fill in a survey. I know many would like to get quality data from the customer in the experience, however understand that customers who are looking are not keen to share their contact details. Often they use “mickey mouse” details – like Gmail or Hotmail data. Worthless anyhow – why not remove it.

It is very simple to track the user adoption. So why not ask for the details within the application and enrich your CRM or prospect database for follow-up using these. Often you will have to ask for a company, a full name – and there you go.

Last but not least. For complex scenarios, I mentioned that these trials would probably be used for some kind of POC. Yes – you do not want to offer free service to the public. Instead of making it a ridgig process – contacting a phone number, tampering data or filling a form – why no use the steps to know your customer. Extensions should be part of the journey and should connect the customer to you.

Think smart about your trial experience. A/B testing is the secret to success. And remember – simplicity is king so adjust.

UPDATE: Please see the article on tools to help setup your structure around trial management.