You can view and download our Research Management System here. It’s free with an optional donation to charity.

In a prior post on learning cultures, we wrote about how the short-term time and energy cost of sharing information with team members can prevent people from sharing their insights and stop firms realising long-term team-learning benefits:

In our experience, the biggest enemy to Level II Learning is marginal thinking (first coined by Clayton Christensen). As humans, we naturally focus on the short-term marginal cost-benefit of our choices while failing to comprehend the long-term total cost-benefit. For example, if I eat this piece of cake, I will get great marginal benefit today (enjoyment) with only a small marginal cost (health reduction). We can rarely conceptualise the total cost-benefit — in this example, that one’s long-term health may be put at risk if this same decision is made too frequently. There are many aspects in life where this affects us. In the work place, it often rears its head when considering any form of change. The cost of change is often large in the short-term, while the benefit is often small in the short-term and becomes very large only in the long-term. The marginal cost-benefit of change is almost always negative, while the total cost-benefit can be massively positive. It is natural then that organisations often avoid change. Indeed, the recommended change is often judged against the unachievable (i.e. perfection) rather than against the imperfect and often inferior “status quo”. As such, status quo gets an easy ride and becomes the greatest adversary of change. In reality, change only needs to be superior than the status quo and we should judge the status quo just as harshly as we judge change. This applies to Level II Learning — taking the extra time to share individual learnings with the team to improve process has a high marginal cost (i.e. time and effort) and while the benefit is often very large, it is realised only in the long-term.

A well-designed research management system can reduce these friction costs, making it easier for a learning culture to flourish. A great system is necessary even for individual investors — we often need to share information with our future selves and this carries the same cost-benefit trade-off. The system should also transcend the investment team. Efficient information sharing and a strong organisational learning culture is essential for all functions.

In building a research management system for our investment process and for my prior firm, it became clear that it did not make sense for the research management system to be separate from other things like the customer relationship management (”CRM”) management system, the investment memos, the contacts database, the work flow management system etc. Firstly, integrating these into one system reduces and centralises the application touch points. Secondly, and more importantly, all of these systems are deeply interconnected so they need to be linked. We therefore set out to build something broader than a research management system. The end product is akin to Wikipedia with customised templates that map to our investment frameworks. In this post, we discuss the key problems we set out to solve and the criteria you should consider when designing or adopting a system.

We’ve used Microsoft OneNote extensively and it is not fit for purpose. We’ve trialled Evernote and the conclusion was the same. Both lack the power of databases and therefore lack the ability to properly sort, filter, and link information sets. Both inevitably result in disparate and inconsistent information. We regretted ever using OneNote due to the high switching costs involved in moving platforms. We’ve looked at multiple SaaS-based investment research systems as well as Salesforce and HubSpot. Typically, these solutions are unnecessarily expensive and they never quite do everything you need. We landed on building our own system in Notion (think Evernote combined with the database structure of Microsoft Excel). We have made it available to view and download for free in order to help other like-minded investors. See here if you are interested.

Problems to Solve

We list below some of the key problems we set out to solve in building our research management system. The inter-related nature of these problems means that they are better solved with an integrated system rather than a set of modular ones. The system built solves most of the problems discussed.

• Notes: The first starting point for us was to efficiently record meeting notes (both internal and external) with the ability to add metadata like dates, attendees, topics, companies etc and for these meeting notes to synthesised in real-time into the relevant investment framework. These notes then needed to be easily shared among the team so that everyone could access what everyone else was working on. At the same time, one should be able to enter and share other notes like thought pieces, general ideas, book summaries, investment committee notes, or anything else that is useful to record. It’s always better to send your own insights and ideas from say an article rather than just forwarding the article to the team.
• Company Research, Industry Research, Thematic Research: The second starting point for us was to record other aspects of research such as our own analysis or snippets of readings etc in a way that enabled us to organise and synthesise the work into our investment frameworks as we went. Housing this in the same system as our notes allows the research to be linked so that its easy to see all the meetings and other notes related to the company we are working on at any given time. Similarly, the research can be linked to the investment decision log so that all the decisions made relating to the present company can be easily seen. The same goes for people related to the company (i.e. the contacts database) etc.

Microsoft Word is suited to a linear work process. That is, you start at the beginning, work down the page until you get to the end. In contrast, investment research is often non-linear. It jumps back and forth depending on the focus areas and scheduling. A Wikipedia-style system is better suited to a non-linear process where each element of your investment frameworks can be worked on in a modular way with a synthesis that links to each of these modular elements. Dennis Hong of ShawSpring Partners writes eloquently on non-linear research in his investment letters.

• Investment Memos: There were three aspects to solve for the investment memos. First, we wanted them to be directly linked to the research so that one could click through from the memo to the supporting work as needed. The end result we had in mind was more like a Wikipedia page than a standalone document. Second, we wanted the document to be a living and breathing thing that always represented the latest thinking. Word/PDF investment memos often become old and stale and rarely represent the latest thinking over time. Third, we wanted to easily grant access to the memos to 3rd parties such as clients.
• Decision Log: The importance of recording all decisions in a detailed decision log will be the subject of a future post. In short, our memories quickly become altered and hindsight-bias inevitably creeps in. As investors, we need to be able to look back and know why we made certain decisions. We needed a place to efficiently record our decisions but to also link them to the associated companies and people.
• Investment Frameworks: We spend a lot of time researching and adapting our investment frameworks. We needed a place to record this in a similar manner to our company research and to be able to link the supporting examples directly to the related company research.
• Contacts and Firms Databases: The obvious use case here was to efficiently record the contact information of the people and firms we have met. We then needed to be able to link these people to the relevant meeting notes, companies, industries, etc so that we could do things like view the meeting notes from every meeting we have ever had with a particular person. Finally, we needed to store the diligence performed on certain people. This could be company management for direct investors, or the portfolio managers and team of 3rd party investment firms for allocators.
• Team Profiles: For larger teams you may even want a database for internal team member profiles (e.g. areas of expertise, companies diligenced at prior firms etc). This helps individuals in the firm know who they can reach out to for help and to avoid duplication of effort.
• CRM System: Most investors will agree that managing the process of meeting with and following up with company management, industry experts etc is not dissimilar to the requirements of managing client meetings and relationships. It therefore makes sense for the same platform to do both.
• Client Letters: The letters we write to our clients are a showcase of our investment thinking and research. However, as with investment memos, when they are sent as standalone PDF documents, they often become stale. Rather than emailing PDF documents, we wanted to give clients access to a Wikipedia-style knowledge centre that is inter-linked and constantly updated with the latest thinking. Indeed, this website is created and hosted within our system. Most of you probably read the Stratechery blog and it is easy to see how good the user experience is when reading a product like that versus a PDF.
• Data Sources: Over time we have come across many valuable company and industry data sources. After a while, it became hard to remember what they were and what they were related to. We needed something to easily track, record, and catalogue these and link them to the relevant companies, industries, and people.
• Workflow Management: This started with the need to centralise the firm’s priorities so that tasks can be easily assigned and monitored, and to do this in a way where they can be easily and automatically re-arranged into different views. For example, management may want to view them in date order or in status order while individuals may just want to see what is assigned to them. This needed to be done in a way where the lists were not duplicated. However, we also needed to link the priorities to the companies, industries, and people involved. Housing this in the same system as the research enables this.
• Tracking Contributions to Team Learning: All of the above are more tangible problems to solve. An intangible problem was the need to monitor and track contributions to building an effective learning organisation. Part of this is the ability to monitor individual contributions to the sharing of information among the team (both quantity and quality). In an ideal world it would also monitor and track the other side of information sharing, i.e. who has read what and when. The system built does a good job on the former but only partly solves the latter. It is possible to see who has read a page and when but not in a systematic way.

Key Functional Criteria

There are various functional criteria in any research management system that should be considered when evaluating different options. We list several below:

Non-Negotiables

• Easy to Use: The starting point for any system should be that it is easy to use. This means that it should be quick and efficient to enter, recall, and share information. It should also be intuitive and easy for team members to learn, even for the most technically-challenged. This helps reduce the inevitable “adoption hurdle” and the natural organisational resistance to change. If you don’t have the luxury of dictating which system will be used, you need to create a coalition of support. In our experience, it helps your cause to first get the most technically-challenged on board by taking the time to explain the long-term benefits to them and that you need their help to encourage others.
• Can be customised and expanded: Your system should fit your own investment process. For example, your meeting notes and investment memo formats should map to your own investment frameworks. Further, you will naturally find new use cases for your system over time. It therefore needs to be expandable with new custom metadata categories and new databases / content. Out-of-the-box solutions are not only expensive, they rarely solve all of your needs. Further, unless you are a very large customer, you will have little influence over the product roadmap and so they lack the required customisation. This means opting for a system that can be easily customised on your own without the use of code.
• Metadata: Easy information recall, sorting, and filtering requires the ability to attach useable metadata to the information such as dates, names, companies, topics etc. This means the system needs to be built on top of a database structure — something that products like Word, OneNote, Evernote etc lack. Attaching metadata also allows for easy automated restructuring of the information views (e.g. pre-set filters or converting a table view into a calendar etc). Not only is this convenient, it helps avoid the duplication of information. For example, if you have a set of team priorities, you would rather filter this view to see individual priorities rather than have a separate and duplicate list of the priorities by individual. You can even have a separately located view of the same database (e.g. your individual priorities view on a different page to the team priorities but they remain one and the same database).
• Database Linking: The ability to link databases (e.g. linking the notes database to the contacts database) means that the subject for one database can become metadata for another database. This allows you to do things like viewing all the meetings you have had with a company or person while on the relevant company or contact page. It also helps keep the databases consistent — changes made in one database are automatically made in the linked database. For example, if a company changes its name, you can edit the company name and it should automatically flow through to any reference / link to the company. These database-links also allow you to easily click through to another database. For example, if you want to view the background of a person listed as an attendee on your meeting note, you can easily click on their name to open this up and jump into the contacts database without separately navigating there from the sidebar. That is, it makes it easier and more intuitive to use.
• Access from anywhere on any device: This means cloud-based with a great desktop and mobile product.
• Concurrent work: Multiple people must be able to work and update the system at the same time, including in the same page without ever creating a duplicate document. As most will attest, while this is theoretically possible in Word and Excel, it usually works terribly in practice. It often means waiting until people have exited a document, or ending up with multiple non-concurrent versions of the document.
• Contribution Tracking: The system should allow tracking of individual contributions. This helps monitor adoption and to inform the performance review process with respect to general contributions to the organisational learning culture.
• Provisioning access internally and externally: The provisioning of access down to the page level should be possible for both internal team members and external parties, including external parties who do not have an account on the platform used. As mentioned above, even this website is created and hosted in our system. We can then easily publish the public content to the web.
• Security: The system should have all the latest security protocols like single-sign on, two-factor authentication, encrypted data etc.
• Others:
• Version control and backup
• Easy import and export of content to facilitate transfers from / to other systems
• Note: Be aware of Microsoft OneNote…they make it very difficult to transition away from the product preferring to capture customers with switching costs rather than customer satisfaction.
• Product has critical mass in the broader market for confidence the service is here to stay
• Links easily to your document management system

Nice-to-Haves

• Cost: This is a nice-to-have. It is a mission critical system for an investment firm and therefore the willingness to spend should be high. That said, Notion is a very cheap product at $20 per user per month for their enterprise service.
• Looks Nice: A great functioning system will be used very frequently and can also be used to provide information to third parties. It should therefore have a nice look and feel to it. In our experience, platforms with limited formatting and pre-designed templates lead to a cleaner look as it forces team members to all use the same styles (you should never have to spend time worrying about why the spacing is off in a paragraph the way you often do in Microsoft Word). OneNote inevitably ends up looking like a dropped plate of spaghetti.
• Content Syncing: A useful feature in Notion is that you can copy a paragraph into another page and actually have them synced so that a change in one is made in the other. This helps reduce duplicate information and to increase consistency so you don’t have to update paragraphs in two separate documents.
• Connects with workflow automation tools: This just makes life easier. For example, when we enter a contact into our phones, the workflow automation we have set up automatically enters this contact into our Notion contact database.
• Others:
• Notifications to team members of new content
• Tracking who has read things and when
• Can serve as the web publishing and hosting tool with custom domains
• Automated alerts for when you should set up meetings with certain people (e.g. clients)
• Integrated with Elon Musk’s Neuralink…maybe one day 🤣

How does Notion Stack Up?

As discussed above, we have built our system in Notion. The below table analyses Notion as a platform to use in building your own system.