A little word of warning: This newsletter is a bit longer than usual. It's a very complex topic so I wanted to cover it more deeply. Feel free to reply to this mail to let me know what you think of it. 

Let's dive in with a little game...

Finding a Leader for the Data Leader

The game goes something like this: I throw some common job titles on the table and we assign a reporting line to them together. Here we go:

→ Head of Performance Marketing -> CMO. Easy peasy lemon squeezy 🍋

→ Head of Engineering -> CTO. We’re getting warmed up here.

→ Head of Controlling -> CFO. This game starts to become a little boring.

→ Head of Data & Analytics -> … -> … *scratching head* -> CTO?? Wait… -> CFO??   

If you look at 5 different companies, chances are that the data leader reports to 5 different C-Level roles. 

In fact, I ran a little survey a while ago that proves my point:

How come it is so hard for data folks to find a “home”? 

So many options

Let’s have a look at some of the possible leaders for the data leader. I want to mention here that I have most experience working with Venture backed scale-ups and PE-backed companies. 

So, I will be focusing on these but most of the points I'm making are also valid for larger companies).

Here is our list of options:

• CEO
• CMO
• CFO
• CTO
• CPO

Some people might complain and demand to add a Chief Data Officer (CDO) to the list. But since we are talking about scale-ups here, I will ignore this now. Having someone with a CDO title in scale-up is like wearing a tux to a BBQ. 

There is one more C-level role that deserves a special mention: the COO. In my experience, I've only encountered one startup that had a COO before reaching Series-B funding stage. In retrospect, I believe it may have been too early to have that role in place. I understand that I’m venturing out onto thin ice here as some may argue that a CFO is also too early for a pre-Series B startup. But from what I've observed, a CFO is typically brought on before a COO. Therefore, I'm going to draw the line at the COO role.  

We are now left with five C-level roles which is still excessive for a pre-Series B company. Therefore, when discussing roles, it's important to note that I'm referring to the responsibilities and tasks associated with them, rather than specific individuals. It's common for founders to take on multiple roles throughout the life of a company.

Let’s decide HOW to choose a leader before deciding WHO to choose as a leader

Which criteria should we look at when deciding who the data lead in our company should report to?

First of all, what does “reporting to” actually mean? 

We need to distinguish between functional leadership and disciplinary leadership. Our Data Leader needs someone to give him/her purpose, a mission and goals.  

Who has the pain? 😷

One of the practices I always preach in every data project is to start looking at the problem before looking at the solution. So, instead of thinking about who our data leader should report to (solution), we can think about which potential leader for our Data Leader needs the data leader's services the most (who has the problem). 

Who is most qualified? 🛠

Data leaders in early-stage companies can easily become overwhelmed by tasks that should not fall within their area of responsibility. These tasks, such as fixing tracking code in the webshop, creating workarounds due to a lack of ERP systems, or spending countless hours automating investor reports with numerous metrics from various applications, can lead to inefficiency and frustration for all parties involved. To avoid this, it's crucial to have a leader who knows how to effectively utilize the strengths of the data leader for the benefit of the company.

Who has the bandwidth? ⏳

Let’s face it: no scale-up Founder has any extra time to deal with this pesky data topic. But someone has to take one for the team and deal with it. Maybe not the one who has ten direct reports already.

What’s the company’s maturity? 👵🏻

The data needs of a company that has recently raised Series-A funding are vastly different from those of a mature company. As the company grows, priorities will change, and the number of C-level roles will likely increase. With this growth, the number of potential leaders for our Data Venus also increases.

What’s the company’s type? 📚

The data challenges faced by a marketing-driven consumer internet company with hundreds of thousands of customers, website visitors and app users are vastly different from those of a product-led B2B enterprise SaaS company with only a few dozen customers. Also, it's not uncommon for a marketing-driven consumer internet company to pivot into a product-led B2B enterprise SaaS company. This is why it's typically too early to establish a dedicated data leader in the pre-Series A stage.

Data Team's operating model ⭕️

It's important to differentiate between the data function and analytics function. The degree of centralization / decentralization of these two functions have a large impact on the right reporting lines. As many of you know, I am a huge fan of hub-and-spoke models. I have seen most success where at least the analytics function is decentralized into business domains and I've also seen success where some part of the last mile analytics engineering is decentralized. However, I usually don't recommend to start like that in scale-ups. I think there should be some business super users who can self serve to some degree (e.g. using Google Sheets sitting on datamarts in BigQuery) but decentalized analysts are usually overkill until well past Series-C. So, I will not look into this point in detail in the following evaluation.

Choosing the lesser of all evils

With the list of our selection criteria in hand, we can now evaluate potential candidates for the role of leading our Data Leader. 

Let's take a closer look at each candidate.

CEO

😷 Pain: The leader of leaders sees things from a helicopter perspective and may lack in-depth knowledge of the different functional areas. Even hands-on CEOs may not be aware of specific data-related pain points until they are brought to their attention by other C-level executives who complain about a lack of data needed for decision-making.

🛠 Qualification: Without deep nitty-gritty knowledge of a specific functional area, it can be difficult for the CEO to provide direction for the Data Leader. This can cause the Data Leader to wander aimlessly. If a Data Leader in an early-stage company reports to the CEO, it's important that the Data Leader is skilled in self-organization. On the other hand, the CEO's oversight of all functional areas of the business makes them a suitable candidate to ensure that the Data Leader prioritizes work that addresses the most critical challenges for the company's success.

⏳ Bandwidth: This is often a massive problem. Since all the other C-Level folks already report to the CEO, there is no additional bandwidth to add a direct reporting line to the Data Leader. Having more than 7 direct reports is rarely practical. A possible setup exists when different co-founders share the CEO role and also take on functional roles such as CTO or CMO at the same time. 

👵🏻 Company Maturity and Type: The data needs of an Enterprise SaaS startup are vastly different from those of a Gen Z dating app, and the data needs of a Series-A Gen Z dating app differ from those of a Series-C Gen Z dating app. It's important to recognize that the reporting line between the Data Leader and their supervisor does not have to be set in stone. However, for mature companies with a large customer base where identifying and retaining profitable customers is crucial to business success and the complexity of managing such a large number of customers is high, it's essential to have the data function report to the highest-level executive.

CMO

😷 Pain: If you were to take a CEO, CMO, CFO, and CTO from the early 2000s, push them into a time capsule and send them to 2025, the CMO would likely experience the most significant shock. The role of the CMO has changed the most in terms of their need for data. Today, CMOs require accessible and reliable data to make important business decisions more than ever before.

🛠 Qualification: CMOs have become well-versed in technology and are often responsible for managing larger technology budgets than CTOs. This makes them well-suited to lead data initiatives. However, unlike CEOs, CMOs may not always remain neutral. There is a risk that data structures and projects will be tailored to the needs of marketing, leaving other C-level executives in the dark.

⏳ Bandwidth: Particularly in the early stages of a company, CMOs typically have fewer direct reports than the CEO, giving them an advantage.

👵🏻 Company Maturity and Type: When a company has just raised its Series-A funding, the main focus is on growth. However, the strategies for achieving growth can vary greatly depending on the company's maturity and industry. For example, our Gen Z dating app may initially focus on acquisition and improving the conversion funnel, while later shifting towards increasing the lifetime value of existing customers. In both cases, there is a strong need to accurately measure and optimize customer acquisition costs (CAC) and lifetime values (CLV), making this type of company a good candidate for having a data leader report to the CMO. On the other hand, companies that follow a product-led-growth strategy (typically SaaS companies) require a different approach. These companies focus on converting free users into paying customers and require a different set of data to do so. While marketing is still important for driving users into the funnel, the more complex data needs lie within the Product department.

CFO

😷 Pain: I don’t envy CFOs in scale-ups. They really have a tough life. One of the reasons for this is the need to have a broad understanding of a large number of key performance indicators (KPIs) rather than a deep understanding of a few. For example, CFOs need to know their margins and inventory value, but depending on how their role is defined and interpreted they often don't need to know margins per customer segment or inventory value per SKU. Unfortunately, data teams in early stage companies are not well equipped to provide a large number of KPIs across all functional areas of the business. They excel at providing in-depth insight into a few KPIs. Another issue CFOs often face is that data teams are not always the best department to fix their pain points. For example, incorrect inventory numbers are often caused by a lack of proper systems (which is the IT department's responsibility) or a lack of proper processes (which is the responsibility of the functional departments such as Operations or Warehousing).

🛠 Qualification: Unfortunately, during the course of my career I have seen many CFOs misusing Data Warehouses for accounting, compliance, investor reporting and other use cases that Data Warehouses are not designed for. I also met CFOs who were extremely uncomfortable with the small inaccuracies that are unavoidable in data teams in early stage companies. Another issue that I see with CFOs in (tech) startups is that they are often the least technical C-level person. Lastly, due to the fact that the CFO requires high level KPIs across all business units I have seen them struggle to focus data teams on getting one use case right first. This led to the data teams getting nothing right. If CFOs can overcome these shortcomings, they are suitable candidates for leading the Data Leader due to the fact that they “own” all company-relevant (top-level) KPIs. 

⏳ Bandwidth: Particularly in the early stages of a company, CFOs - just like CMOs - typically have fewer direct reports than the CEO, giving them an advantage.

👵🏻 Company Maturity and Type: If a CFO doesn’t disqualify due to any of the criteria mentioned in the qualification section, the CFO is a suitable candidate irrespective of company maturity and type. 

CTO

😷 Pain: The CTO is usually the one person on the list who doesn’t feel a lot of pain if there is a lack of accessible and trustworthy data to drive decision-making. This can be an advantage, as the CTO won’t push his own agenda when leading the data team.

🛠 Qualification: This is a difficult one. As mentioned above, a big advantage of the CTO is his neutrality when it comes to prioritizing the roadmap for the data team. Being the most technical C-level executive (unless the CMO starts out-nerding the CTO) you might also think that the CTO is a natural fit for the data team. Very often this is exactly what happens. “Data and IT? They all do stuff with computers and numbers right? So, let’s put them together”. The problem is, though, that building operational applications powering a web-app or mobile-app or a shop-backend is completely different from building analytical systems for decision making. This is often overlooked. The consequence is a CTO who is overwhelmed by the additional overhead the data team requires. Since it’s not his home turf and since there is an immediate revenue loss when the website is down but not when the data warehouse is down, the CTO ends up neglecting the data team. Also, the CTO often does not have the deep business understanding that is required to guide the Data Leader. However, if IT and data are under one roof, it can lead to better development of data contracts and handover points, avoiding issues with data pipelines in the data warehouse.

⏳ Bandwidth: Especially in the early stages of a company the CTO personally manages all engineers without an additional level of management in between. Also, it is not uncommon that the CTO builds large parts of the application by himself. When any operational application is down, all hell breaks loose, investors call every 5 minutes and customer complaints start piling up in the customer service team. Not a good environment for our young data leader to grow. 

👵🏻 Company Maturity and Type: The decision whether the data lead should report to the CTO has not so much to do with the company maturity and type but rather with the qualification and bandwidth.

CPO

😷 Pain: Not many people usually think of the CPO as a potential leader for the Data Leader. I think that this is shortsighted. We talked about product-led growth companies before. These type of companies are often underserved with data. They often have great product analytics tools such as Mixpanel, Amplitude but without connecting that data to the backend databases they are lacking a holistic view of their customers.

🛠 Qualification: Product Managers often lack even very basic data analytics skills in the same way as data people often lack very basic product management skills (such as listening to customers). A combination of solid data skills and the product manager’s ability to talk and listen to stakeholders is a killer combination. A nice side-effect when the data team reports to the CPO is that feature development, tracking and data is under one roof. Less surprises for the data team when new data structures and events that need to be implemented into the data model start popping up.

⏳ Bandwidth: This depends a little bit on how the CPO interprets her role. A very hands-on CPO will probably struggle with the context switch between data and product but a CPO with a strong independent team of product managers can definitely thrive at being a great leader for the data team.

👵🏻 Company Maturity and Type: As mentioned already, tying the data team to the CPO works best in product-led companies and also works better in the earlier stages (say between Series-A and Series-B). In later stages there is the same danger as with the CMO that other business units’ data needs are neglected while the CPO might push her own agenda.

Bottom Line

Finding the right leader for the Data Leader is a topic that’s very close to my heart. 

Over the past few months, I have been thinking and talking a lot about this topic. 

During one of those conversations someone told me this:

• Data Leader reports to the CEO -> company cares a lot about data
• Data Leader reports to the COO -> company cares about data
• Data Leader reports to the CIO -> company does not care so much about data
• Data Leader does not report to any CxO -> company does not give a 💩about data

While this is a very simplified way of looking at the problem, there is some truth behind it.  

At the end of the day, the perfect leader for the data leader will be different at every company. But taking into account who has the pain, who is qualified, who has the bandwidth and the company’s maturity and type can help make a confident decision.

That's it. I hope you don't mind the longer read today.

Cheers,

Sebastian

P.S.: Over the last 18 years, I've built data teams that reported into all 5 functions. I am sharing all my frameworks, strategies and blueprints in my ​Masterclass "From Dashboard Factory to Strategic Partner."