Over the past years I have been focused on understanding how data can help a startup reach the growth phase, and support growth and scalability once it got there. It is a trend nowadays when every startup speaks of data-driven processes, but for many as I discovered this is merely plug and play external tools into their regular channels and dive into predefined reports. Whenever there is a challenge to customize those inputs, there is a default to Excels, Spreadsheets and lots of manual work where teams, founders and providers try to create, with a substantial time waste, a customized data room and reporting system.

Building a data-driven culture within a small (or large) team means ensuring everyone has access to data in a simple way that makes sense for their job and goals. This implies spending less time collecting and processing data, and more time analyzing it. You would be amazed at how ingrained data-driven decisions become in your organization, once people stop collecting and reformatting data, and instead are able to analyze it whenever they need to.

For founders this also stays true. For board meetings, strategic or day to day decisions, or fundraising, founders spend an enormous amount of time simply collecting data (or having their team collect it for them), transform it, process it, perform calculations and then as an end step creating visualizations that can be shared with the rest of the stakeholders. 

I was there, and I know the time wasted to get a pitch deck updated just to realize at the end the data did not help much with my efforts. And I wondered every time, how come I did not see this earlier. And a few months later, the same process would happen again, and a new version of a data room would come out, showing more surprises. 

Select the metrics your data room needs

Many founders go directly to the basic metrics that investors ask for, but this may not always be the only data you need. Your business is different from most of the startups out there, and you need some customization. One time, we ended up developing a 70 slide business overview business performance report. At the department level we created 20 slides sales decision performance report, a 3 slide customer experience report and at marketing level we had 3 reports, roughly 5 slides each. The beauty of it was that they were all automatic and fully customized, while maintaining information that was still usable for a fundraising pitch deck and data room.

For example the business overview holds sales funnels that are taken from the sales reporting, lead generation costs per channel from the marketing reporting, and churn and MRR data from the customer experience reporting. 

What we can surely say is that some of these metrics are similar at market / industry level, and they should be standardized as much as possible, calculated only once, and shared across each business unit as much as possible. While your own metrics / reports are surely different than what I will detail below, the logic finding where the data resides is similar across the board.

Financial Data

The single most important category most likely, as financial metrics are always required at investor meetings, and they should be present at every weekly / monthly management meeting. These metrics should be calculated over time, and historical data is quite important. Examples of calculating these metrics:

  • MRR / ARR. The source is the list of transactions performed since you started charging customers that holds: clientID, date, amount, frequency. This is enough for the basic MRR calculations. 
  • Customer Lifetime Value can be tracked from a transaction table. I like to calculate customer lifetime value at customer level, then I can use this value in cohorts regardless of the cohort type / period. Same for customer lifetime. For eCommerce, I also calculate the total number of transactions performed, which allows for flexible calculations for the average number of transactions over any period of time. If I choose to show a report over quarter or months or years, the data will suffice the purpose.
  • Cash runway can be tracked by crossing your accounting system’s balance data, with your budgets data, usually a mix between accounting stored info and a spreadsheet, and sales forecast data which should come from your CRM’s sales and clients funnels. 

Product Data 

In most digital products these can be seen as behavioral metrics. Tracking them is key to understanding product usage growth, as well as acquisition, activation and retention of users, which can give more insight into predicting future growth. 

  • DAU / WAU  are event driven metrics, and you would do well to implement an event tracking system for your product just about now if you want to track these. This data should be measured inside the product, and sent to an external tool (like Amplitude or Mixpanel) for visualization. ClientID plays an important role in mapping these metrics correctly, as is clearly defining what is a “valuable” activity for your users / clients.
  • Retention depending on the business model, this can be more complex in terms of properly tracking it. If you are a Saas business, you may want to look at retaining clients as much as retaining users. Client retention is in itself a financial metric, and should be calculated based on the payment transactions tables, while user retention is more product focused, and should be tracked with key events (core feature usage) based on the same event tracking system and points from DAU/WAU.
  • Churn similar to retention, revenue churn should be calculated based on financial transaction history, while user churn based on basic events (e.g. login) and core events inside your product. 

Sales Metrics

For most B2B and Saas sales, metrics will be centered around funnel, lead to client conversions, sales duration, contract amounts, sales rep performance, geography. In the B2C businesses the funnels are built more on the marketing side, while sales per se will focus on cart value, number of purchases, etc.

  • Qualified Leads, in case of a B2B product, the CRM should be your main source of data for sales metrics, and qualified leads should all be registered there. In a B2B or B2C Saas or eCommerce product, this information can be tracked inside the product using events. A qualified lead for an eCommerce platform can be a user that performed a signup and followed that by several product view pages or even added to cart within a predefined period of time.
  • Pipeline Value, can be seen directly in the CRM system. The reason I recalculate this metric is I can enrich it with custom fields, and cross it with marketing data to really dive into it. A CRM will let you do some reporting configurations, but most of the time it will not be very flexible when it comes to your own custom fields (business subcategory, client’s number of employees, main challenges, etc). When collecting a lead for Saas for example, I ask the person to fill in his/hers main challenges. Having this information, and the attribution, I can now break down which issue is most common with sales qualified leads that convert to clients, and come from my Google Adwords campaigns for example. 
  • Conversion Rate is again something most CRMs offer by default at company and sales rep level. While it is usually enough, having access to the data allows for an analysis of the conversion rate per sales rep at a granular level, for e.g. it can be broken also by business type / account value / business size / business challenge. It can show you rep A can be highly efficient with small accounts that have problem X while rep B is highly efficient with large accounts that have problem Z.

Marketing Metrics

Especially easy to automate and track for digital marketing. These metrics should focus on channel performance, channel value, channel scalability, channel ROI mainly. 

  • Customer Acquisition Cost: based on attribution and marketing spend. In order to calculate correctly, each user / client needs to be stored with the original attribution data. The CRM / database should be configured to retain the original channel information.
  • Lead to Customer Conversion Rate: can be calculated from the CRM data, or tables in our database for leads and customers, where we store at least attribution, lead generation / signup date, free to customer conversion date. 
  • CTR: provided usually by Google Analytics, but best to calculate it dynamically using actual number of pageviews from a Google Analytics export. If we look at CTR for an ad, then we can cross the view ad data from the ad platform with the pageview from our analytics or screen view data from our mobile app. 
  • Traffic to lead ratio: we can cross this based on date, by using the Google Analytics sessions/users data and the CRMs leads data crossing the traffic date with the lead creation date in the CRM. If we stored attribution in the CRM we can then breakdown this data at channel level. 

In order to create an automated channel value or ROI report you need to make sure that every transaction is attributed correctly. Attribution plays a key role in building a strong automated reporting system, and it should always be a priority when launching marketing campaigns, saving transactions in a database, or leads in a CRM. Channel attribution should be preserved all the way from lead generation to the final stage in the funnel.

A couple of thoughts: 

  • Saas / B2B save your UTMs / tracking fields in custom fields inside your CRM. Make sure you have a set of initial attribution fields, and a set of fields for last touch attribution. Always update automatically the last touch attribution fields, while the initial attribution only update those the first time a lead is generated.
  • For eCommerce / B2C make sure you save the original attribution at user level, and each transaction / activation / retention event at first touch / last touch model if possible. This data will allow you to understand which user acquisition channels work best, and what channels to use to keep your users active.
  • Mobile data attribution will always be more work to store it and access it properly but can be passed on into the app events tracking, and thus embedded into the data strategy. 

Operations Metrics

These would be centered around HR, fixed and variable non-marketing and non-sales costs, customer response times and tickets, customer incidents, etc. 

  • Runway is a simple cross between your forecasted revenue, calculated for a recurring revenue business for example based on the active account’s MRR, and crossing this data with forecasted budgets / spending that can either be extrapolated from your accounting systems past 12 months data, or a manual input of generic category cost (for small teams might be easier / faster).
  • Overhead cost per client – founders rarely calculate this and one of the reasons is the data is not easily accessible. You can store support tickets associated with an accountID and the employeeID who handled the request. You have a transactions table with clientIDs which can be mapped to an individual user accountID. You also have a table with the HR costs per employeeID. This all can be crossed to calculate not only averages of overhead but actual overhead at account level. Crossing this with business type and size, and you can understand which client categories are profitable, and which are actually making you lose money.   

The data required for these metrics should be extracted from 

  • accounting system
  • customer support system
  • software development tracking software

Applicable to all areas, we can design a company wide Target spreadsheet, where we map out the targets we want to achieve. From sales revenue, to cost per lead, to number of new hires, employee retention, tickets resolved, time to resolve a ticket all can be easily set up in a spreadsheet that is accessible to each team manager. Once defined, these target metrics can be brought into the data model and linked with a calendar table to track performance and target achievement rates.

What tools can you use to build our automated data room?

There is a wide range of tools out there that can provide insights into some of these metrics, and that can build reports on top of the data a startup collects. The challenge I have many times is most of them are focused on one area: Marketing, Sales, Product, Financial. Some of these tools manage to go in depth in two areas – like marketing and sales. But few of them give a complete overview of the business and can cross the data from different areas easily. 

Out of the box 

For marketing and sales there is a wide range of tools, free and paid, that take the work out of building reports. Google DataStudio, Cyfe, Databox, Geckoboard, Octoboard are just a few that help out building automated reporting capabilities.

On the financial and operations side tools like ChartMogul can provide some insights for recurring revenue reporting, analytics & metrics. Freshbooks, Xero though mainly accounting tools, can also provide some integration and reporting. Similar Stripe for example has the transaction data available through APIs.

Visible.vc aims to integrate a little bit more the financial, marketing, sales and some operational side. 

For product data we have Amplitude, Mixpanel, Looker, Rakam. All good in their respective fields, some more affordable than others. 

While I love good products (disclaimer, I have been a user and supporter of Amplitude for a few years now) and I strongly recommend using some of the out of the box tools for fast deployment of data reports at department level, many of these tools do not always match the boardroom or fundraising needs. With several of these tools in my kit, I still had to build custom reports for investors, and even more, cross data between silos in order to bring the full picture in the meeting / due diligence process. 

Integrating Tools 

My preferred approach is integration of data sources, into a data transformation & visualization platform that supports my data room. Each of our systems from our social ads platforms, to our CRM, emailing and accounting systems, to the payment processor provide an API or integration with a database-like tool (Airtable, Google Spreadsheets, Excel online). 

Once the data is stored, tools like PowerBI or Tableau – when there is no direct integration to the data source system – easily extracts and transforms data, storing it in a virtual database tables environment where advanced mix and match can be performed.

In the next article, I will dive into practical examples on how to use these tools together to build your automated data collection and reporting system. We will take them for a short ride, and map out how they integrate and how to build the most important reports a startup’s data room may need.