Gut instinct is a good thing, many times you need to rely on your intuition. You should never ignore your gut feeling, but you need proof. It’s all about asking the right questions and concentrating on the key metrics that drive the changes that you are looking for.

Track marketing metrics that are critical to your startup

You are just wasting your time if you measure something that is not attached to a goal. You need to find a metric, that changes the way you behave. The best metric is a ratio or a rate, that is comparable to other periods.

What business are you in?

The best way to think of metrics is to consider how you plan to charge your users. According to startup investor Anu Hariharan, there are 8 main business models, which most startup entrepreneurs fall in:

Saas business model

A SaaS (software-as-a-service) company sells subscription-based licenses for a cloud-hosted software solution in most cases to B2B users. (The cost of the transaction is high.) You charge something monthly for the software you provide. Most startup companies match this category. Its metrics:
MRR (monthly recurring revenue): revenue recognized for recurring services rendered in a given month.
ARR (annual recurring revenue): ARR = MRR*12
MRR-based churn: monthly recurring revenue lost in a specific month / monthly recurring revenue at the beginning of the month.
CAC (cost per acquired customer): cost per customer acquired through paid marketing channels.

Subscription business model

A subscription company sells a product or a service, usually to consumers (B2C market), on a regular basis. (The cost of the transaction is low.) Example: Netflix. Metrics:
MRR (Monthly recurring revenue): revenue recognized for recurring services rendered in a given month.
GR (Growth rate): it is based on monthly recurring revenue. GR = (latest month MRR/first-month MRR)^(1/#of months)-1
Gross user churn: canceled subscriptions in a given period/prior period total customers
CAC: cost per customer acquired through paid marketing channels.

Enterprise business model

An enterprise company sells software or services to other businesses on a single-license basis. Enterprise company works based on a contract. Metrics:
Bookings: sum of the value of all customer contracts
Total customers: total number of unique contracted customers today
Revenue: business results after delivered services.

Transactional business model

A transactional company enables a financial transaction on behalf of a customer and collects a fee. Example: Wise. Metrics:
GTV (Gross transaction volume): total sales or payment volume transacted in a given period
Net revenue: the portion of GTV that the company defines as revenue
User retention: it is a cohort metric; percent of customers who go on to make at least one purchase in month 2.
CAC: cost per customer acquired through paid marketing channels.

Marketplace business model

A marketplace company acts as an intermediary between sellers and buyers, generally collecting a percent of the total transaction value. Example: Airbnb. Metrics:
Gross merchandise value: total sales volume of merchandise in a given period.
Net revenue: the portion of GMV that the company defines as revenue
GR (growth rate): it is based on net revenue. GR = (latest month net revenue / first month net revenue)^(1/#of months)-1
User retention: percent of customers who go on to make at least one purchase in month 2 (this is a cohort metric)
If your business cycle is long, look for indicators of customers’ behavior before purchase.
CAC: cost per customer acquired through paid marketing channels.

e-commerce business model

An e-commerce company sells physical goods online. Generally, e-commerce companies manufacture and inventory those goods. Example: Amazon. Metrics:
Monthly revenue: total revenue in a given month
GR (Growth rate): it is based on revenue. GR = (latest month revenue/first month revenue)^(1/#of months)-1
Gross margin: gross profit / total revenue in a given month
CAC: cost per customer acquired through paid marketing channels.

Hardware business model

A hardware company sells physical devices to B2B or B2C markets. Example: Samsung. Metrics:
Monthly revenue: total revenue in a given month
GR (Growth rate): it is the monthly revenue growth rate between two disparate months. GR = (latest month revenue/first month revenue)^(1/#of months)-1
Gross margin: gross profit in a given month/total revenue in the same month
CAC: cost per customer acquired through paid marketing channels.

Advertising business model

An advertising company offers a free service to consumers and derives revenue entirely or predominantly from advertisers. Example: Reddit. Metrics:
DAU (Daily active users): total number of unique users active in a day.
MAU (Monthly active users): total number of unique users active at least once in the last 28-days.

Startup Metrics. What stage are you in?

However, in the case of a startup, it may be difficult to assess what business you are in, as you are continuously experimenting with offering the right product to the right target customers. Therefore, you need to make step-by-step progress.

At first, you need to know in what stage you are in because there should be one metric to rule them all. Alistair Croll and Benjamin Yoskovitz defined the following stages in their book „Lean Analytics”:

Empathy

Your goal is to drive visitors to your site or generate organic or paid attention.
Startup metrics: traffic, cost per click, email open rate, cost of acquisition.

Stickiness

Your goal is to engage visitors and turn them into prospects or users.
Startup metrics: signups, subscriptions, registrations.

Virality

Your aim here is to convince users to come back again.
Startup metrics: retention rate, engagement, time since the last visit, number of active users, churn.

Revenue

Your goal is to generate money from your business, but it highly depends on your business model (ad clicks, purchases, subscriptions, etc.)
Startup metrics: customer lifetime value, conversion rate.

Scale

Your goal is to find a new market or invite other potential users.
Startup metrics: invites sent, viral cycle time.

The dangers of being too data-driven

There are perils of being too data-driven because there are non-quantifiable driving forces behind objective numbers. Therefore you should collect both quantitative and qualitative proofs.

Quantitative data is objective; it is the numbers we track and measure. It is quite scientific, but it’s hard enough to get the full picture because quality cannot be quantified. For example, you need a qualitative approach to contact your target audience and ask them what problems they’re facing.

However, qualitative data is subjective and not easy to understand, but it is essential to know the internal drivers („WHY”) behind an action („WHAT”). In this case, you are not measuring data numerically, but you are speaking with people instead.

The most important rule is to know your customer. There’s no substitute for engaging with your target audience directly. As for KPIs: you should make early assumptions and set targets for what you think success looks like. But stay flexible, and ready to iterate if needed.

Find a balance between intuition and data-driven decision-making to make the best choices for your business!

Build, measure, learn! Avoid the pitfalls, and don’t jump in without proofs!