An Intro to Startup Operations

Image

Operational requirements vary greatly between Startups, and this means that it's not a matter of simply listing what tools or products to use, or specifying what growth model or customer success model to ascribe to. Consequently we need a framework to understand Startups in general, and a methodology to translate that understanding into practical steps.

This post introduces the concept of modelling Startups as a Production Line, which allows us to use a host of Operations Management techniques to understand and optimise them holistically. Startups must make trade-offs as they grow; the common metrics of CAC and LTV enable us to understand how to balance the Startup, and determine where to dedicate time and resources.

Modelling a Startup as a Production Line

Startups varying greatly in terms of their products, markets, business models etc. However, they all share a host of common elements - for example, some form of marketing or top-of-funnel activity leads into some kind of sales, which results in active users. Active users need to be serviced in some way - perhaps payments need to be made, or customer support is required, or upselling is a possibility.

When we recognise these commonalities, we find that 90+% of Startups can be simplified down to a series of connected processes - just like a production line: -


I'm certainly not saying that every Startup is identical to this - each Startup will vary in its own way and the diagram will be different depending upon the type of operations - but this structure of connected elements can be used to model almost any Startup. For example, a Startup may have 2 inbound marketing channels leading into inbound sales, plus an outbound sales channel - but these are just minor modifications to our model of a Startup as a production line.

Each stage of this production line takes inputs, processes them in some way, and produces outputs. For example, a Sales component may take inputs of MQLs and sales people, and produces an output of Active customers that have signed up for a base subscription. We can represent each block in the following manner, with a set of Inputs, a Function f(), and a set of Outputs - where the Output = f(Inputs): -


This is a great step in modelling a Startup, but first let's illustrate why this is important

The Parable of the Over-Enthusiastic Founder

Imagine a Startup with a young, energetic and overly enthusiastic Founder. The Startup is small, but it has a great product, good staff and a bright future. The Startup needs to grow, and the Founder knows that they need to improve their Revenue in order to get more investment, so they push their Sales staff as hard as possible to bring in more Sales

The Sales staff dutifully do so - meeting customers, hustling, selling deals - and they bring in a load of new Customers. The Founder is happy and praises them all.

A few months later though, and the Startup's Revenue hasn't grown. It's dire. Customers are unhappy, and they're dropping like flies. So the Founder goes to the Customer Support team and berates them - "Sales has done a great job bring in these Customers, and now you're messing everything up by losing them!!!" The Founder is pissed

So What Happened Here?

This is a very simplistic example to illustrate a point, but it's likely that the Sales people have reacted to the pressure placed on them by bringing in a load of unsuitable Customers, who are churning because they were mis-sold the product, weren't the right fit, or didn't realise what they were buying. Customer Support was (hopefully!) doing the best job they could, but was in an impossible situation. From a distance this is easy to diagnose, but I see this sort of behaviour repeatedly in Startups.

Startups far too often focus their efforts on one area and just shift the problem elsewhere because they don't consider the bigger picture

There are myriad examples of this sort of problem: -

  • Salespeople discount prices to close the deal, but kill profitability
  • Marketing targets leads with a high conversion rate to MQLs, only to find that Sales can't close those deals
  • Customer Support win a lapsed Customer back, only to lose them again because they're the wrong fit
  • Customers with poor credit ratings are given credit by Salespeople, only to fail to pay subsequently and create bad debt

The fundamental problem is that it's too easy to push the problem off to another part of the organisation, and without a holistic representation of the whole company, these trade-offs are difficult to see. Ideally we want to find a way to optimise our entire Startup without suffering from enormous trade-offs

Specialisation as a Curse

Unfortunately the trade-offs we've discussed above are exacerbated by high degrees of functional specialisation within Startups. The areas of Marketing, Sales, Accounting, Finance, Customer Support etc. all have their own methodologies, tools, process, and specialists who have devoted their time and careers to a specific area.

There are large barriers between these functional areas - they often work in different tools (e.g. Hubspot // ZenDesk // Intercom), use different terminology for the same things, have specific rituals or practises, and have strong identities tied to their specific functional area. This exacerbates the problem of optimising holistically, as it separates and silos areas from each other.

A Solution

We'd previously considered that a Startup can be modelled as a series of interconnected stages, like a Production Line. Each stage has various Inputs, Outputs, and the relationship between these could be modelled mathematically as some form of linear or non-linear equation.

Hang in here with me on this one....

So we could hypothetically join up each component of our Production Line model, with the Outputs from one stage becoming some of the Inputs to the next stage: -


Keeping hanging in here....

Now we can just formulate an enormous linear/non-linear mathematical equation that represents the whole company and algebraically optimise that equation to optimise our entire Startup!!!

Except....that's incredibly difficult to do, would be really inflexible, and the Startup would probably have changed by the time you manage to do all this work.

A Better Solution - CAC:LTV

We're trying to find a way to understand and optimise our Startup holistically without suffering trade-offs between different functional areas. Modelling the entirety of the Startup mathematically and then optimising it algebraically is one option, but thankfully a better option already exists - CAC:LTV

CAC (Customer Acquisition Cost) and LTV (Customer Lifetime Value) are two common metrics used to measure a Startup, and they have the wonderful property that they effectively cover the whole operations of a Startup: -


All the stages (and costs associated with each stage) that lead up to a Customer becoming Active are covered in CAC. And each stage or function that occurs after the Customer becomes Active is covered in LTV.

Consequently, the ratio of CAC:LTV tells us holistically how the Startup is doing. An ideal ratio of CAC:LTV is 1:>3, and any ratio below 1:2 certainly demands attention. Any changes to operations practises (example, targeting different Leads, or selling by a new channel) will play out in the way these metrics change, and consequently we can predict and subsequently measure the impact of changes holistically.

The ratio of CAC:LTV tells us holistically how a Startup is doing, and is the core way in which we can understand how to optimise a Startup's operations

Furthermore, we can use estimates of how CAC and LTV change with new processes or practises to understand where we should focus our attention. For example, if a Startup has a CAC:LTV ratio of 1:1.5, then their options are either to decrease CAC or increase LTV. They might try to brainstorm new sales channels or methods, but realise that none of these options can meaningfully decrease their CAC - so they're left with needing to increase LTV. That means they need to focus on either the revenue (price) aspect, or on customer life. This might lead them to focus on reducing churn by improving Customer Support, and result in them implementing a complex support and ticketing system to ensure that User issues are dealt with rapidly and effectively.

Analysing CAC:LTV is the main means by which we can understand the operational needs of a Startup, and address where attention and efforts should be focused.

CAC and LTV

CAC and LTV are both fundamentally simple concepts to understand, but practically complex to implement correctly. For this reason we'll give a quick outline here, but will dedicate future posts to explaining both and providing methods for calculating them.

CAC answers "How much did does the average customer cost to acquire?", and is a historical measure

LTV answers "How much value will my startup derive from an average customer over their lifetime?" and is a future estimate based upon observable patterns

The ratio of CAC:LTV therefore indicates trade-offs, for example; between acquiring a customer faster but them yielding less value, or chasing larger accounts that are more difficult and expensive to close, but have a larger deal size. The ratio also helps indicate where focus should be applied; for example, in focusing Product on new features that may increase Revenue, versus improvements to existing features that will increase Retention.

Components of CAC

  • Marketing Spend
    1. Marketing spend on Adwords
    2. Landing page cost
    3. Cost to purchase leads lists
    4. etc. etc.
  • Sales Costs
    1. Sales salaries, commissions, bonuses etc
    2. Overhead for Sales (including office space, computers, etc)
    3. Sales automation tools (e.g. CRM, automated email)
    4. etc. etc.

Components of LTV

  • Value to the Startup of an average customer - on a Gross Profit basis, not Revenue
  • LTV increases with spend
  • LTV is roughly proportional to the life duration of the Customer on your platform
  • LTV decreases with churn (it's better to retain Customers for longer)

These are by no means exhaustive lists of the components of CAC and LTV, but they give a rough overview of what to consider and what affects each. For more details, please see my posts on CAC and LTV methodologies.

Takeaways

Operations varies drastically between Startups; so it's impossible to present a cookie-cutter approach to how to optimise yours. Instead we need to develop a means to understand and optimise any startup holistically, and the first step in that is representing a startup as a production line.

Startups fundamentally need to deal with a number of trade-offs as they iterate around their business and growth models. It is far too easy to create trade-offs that benefit one part of the organisation at the expense of another, and so it is important to consider the business holistically. Modelling the Startup as a Production Line allows us to do this, and opens up an array of Operations Management techniques that can be used.

The "Startup as a Production Line" methodology requires some simplification to be practical, and CAC and LTV are core metrics that provide that simplification and allow us to understand a Startup holistically. However, they are relatively complex metrics that cannot be blindly applied, and require understanding.

We have already published a Lighthearted Intro to CAC-LTV, and subsequent posts will discuss various methodologies for calculating CAC and LTV for different types of companies.

Share this Post:
Posted by Phillip Gales

Phillip is a serial entrepreneur who specialises in Operations, Data and Metrics. He applies AI and Machine Intelligence to old, antiquated and/or forgotten industries that are ripe for disruption.

Phillip holds an MBA from Harvard Business School, and an MEng in Electrical Engineering from the University of Cambridge, specialising in Machine Intelligence.

No comments yet

Leave a Comment