At some point in a startup’s lifecycle, they decide that they need to be ready to go public in 18 months, and a flurry of IPO-readiness activity kicks off. This strategy focuses on a company working on IPO readiness, which has identified a gap in their internal controls for managing access to their users’ data. It’s a company that wants to meaningfully improve their security posture around user data access, but which has had a number of failed security initiatives over the years. Most of those initiatives have failed because they significantly degraded internal workflows for teams like customer support, such that the initial progress was reverted and subverted over time, to little long-term effect. This strategy represents the Chief Information Security Officer’s (CISO) attempt to acknowledge and overcome those historical challenges while meeting their IPO readiness obligations, and–most importantly–doing right by their users. This is an exploratory, draft chapter for a book on engineering strategy that I’m brainstorming in #eng-strategy-book. As such, some of the links go to other draft chapters, both published drafts and very early, unpublished drafts. Reading this document To apply this strategy, start at the top with Policy. To understand the thinking behind this strategy, read sections in reverse order, starting with Explore, then Diagnose and so on. Relative to the default structure, this document has been refactored in two ways to improve readability: first, Operation has been folded into Policy; second, Refine has been embedded in Diagnose. More detail on this structure in Making a readable Engineering Strategy document. Policy & Operations Our new policies, and the mechanisms to operate them are: Controls for accessing user data must be significantly stronger prior to our IPO. Senior leadership, legal, compliance and security have decided that we are not comfortable accepting the status quo of our user data access controls as a public company, and must meaningfully improve the quality of resource-level access controls as part of our pre-IPO readiness efforts. Our Security team is accountable for the exact mechanisms and approach to addressing this risk. We will continue to prioritize a hybrid solution to resource-access controls. This has been our approach thus far, and the fastest available option. Directly expose the log of our resource-level accesses to our users. We will build towards a user-accessible log of all company accesses of user data, and ensure we are comfortable explaining each and every access. In addition, it means that each rationale for access must be comprehensible and reasonable from a user perspective. This is important because it aligns our approach with our users’ perspectives. They will be able to evaluate how we access their data, and make decisions about continuing to use our product based on whether they agree with our use. Good security discussions don’t frame decisions as a compromise between security and usability. We will pursue multi-dimensional tradeoffs to simultaneously improve security and efficiency. Whenever we frame a discussion on trading off between security and utility, it’s a sign that we are having the wrong discussion, and that we should rethink our approach. We will prioritize mechanisms that can both automatically authorize and automatically document the rationale for accesses to customer data. The most obvious example of this is automatically granting access to a customer support agent for users who have an open support ticket assigned to that agent. (And removing that access when that ticket is reassigned or resolved.) Measure progress on percentage of customer data access requests justified by a user-comprehensible, automated rationale. This will anchor our approach on simultaneously improving the security of user data and the usability of our colleagues’ internal tools. If we only expand requirements for accessing customer data, we won’t view this as progress because it’s not automated (and consequently is likely to encourage workarounds as teams try to solve problems quickly). Similarly, if we only improve usability, charts won’t represent this as progress, because we won’t have increased the number of supported requests. As part of this effort, we will create a private channel where the security and compliance team has visibility into all manual rationales for user-data access, and will directly message the manager of any individual who relies on a manual justification for accessing user data. Expire unused roles to move towards principle of least privilege. Today we have a number of roles granted in our role-based access control (RBAC) system to users who do not use the granted permissions. To address that issue, we will automatically remove roles from colleagues after 90 days of not using the role’s permissions. Engineers in an active on-call rotation are the exception to this automated permission pruning. Weekly reviews until we see progress; monthly access reviews in perpetuity. Starting now, there will be a weekly sync between the security engineering team, teams working on customer data access initiatives, and the CISO. This meeting will focus on rapid iteration and problem solving. This is explicitly a forum for ongoing strategy testing, with CISO serving as the meeting’s sponsor, and their Principal Security Engineer serving as the meeting’s guide. It will continue until we have clarity on the path to 100% coverage of user-comprehensible, automated rationales for access to customer data. Separately, we are also starting a monthly review of sampled accesses to customer data to ensure the proper usage and function of the rationale-creation mechanisms we build. This meeting’s goal is to review access rationales for quality and appropriateness, both by reviewing sampled rationales in the short-term, and identifying more automated mechanisms for identifying high-risk accesses to review in the future. Exceptions must be granted in writing by CISO. While our overarching Engineering Strategy states that we follow an advisory architecture process as described in Facilitating Software Architecture, the customer data access policy is an exception and must be explicitly approved, with documentation, by the CISO. Start that process in the #ciso channel. Diagnose We have a strong baseline of role-based access controls (RBAC) and audit logging. However, we have limited mechanisms for ensuring assigned roles follow the principle of least privilege. This is particularly true in cases where individuals change teams or roles over the course of their tenure at the company: some individuals have collected numerous unused roles over five-plus years at the company. Similarly, our audit logs are durable and pervasive, but we have limited proactive mechanisms for identifying anomalous usage. Instead they are typically used to understand what occurred after an incident is identified by other mechanisms. For resource-level access controls, we rely on a hybrid approach between a 3rd-party platform for incoming user requests, and approval mechanisms within our own product. Providing a rationale for access across these two systems requires manual work, and those rationales are later manually reviewed for appropriateness in a batch fashion. There are two major ongoing problems with our current approach to resource-level access controls. First, the teams making requests view them as a burdensome obligation without much benefit to them or on behalf of the user. Second, because the rationale review steps are manual, there is no verifiable evidence of the quality of the review. We’ve found no evidence of misuse of user data. When colleagues do access user data, we have uniformly and consistently found that there is a clear, and reasonable rationale for that access. For example, a ticket in the user support system where the user has raised an issue. However, the quality of our documented rationales is consistently low because it depends on busy people manually copying over significant information many times a day. Because the rationales are of low quality, the verification of these rationales is somewhat arbitrary. From a literal compliance perspective, we do provide rationales and auditing of these rationales, but it’s unclear if the majority of these audits increase the security of our users’ data. Historically, we’ve made significant security investments that caused temporary spikes in our security posture. However, looking at those initiatives a year later, in many cases we see a pattern of increased scrutiny, followed by a gradual repeal or avoidance of the new mechanisms. We have found that most of them involved increased friction for essential work performed by other internal teams. In the natural order of performing work, those teams would subtly subvert the improvements because it interfered with their immediate goals (e.g. supporting customer requests). As such, we have high conviction from our track record that our historical approach can create optical wins internally. We have limited conviction that it can create long-term improvements outside of significant, unlikely internal changes (e.g. colleagues are markedly less busy a year from now than they are today). It seems likely we need a new approach to meaningfully shift our stance on these kinds of problems. Explore Our experience is that best practices around managing internal access to user data are widely available through our networks, and otherwise hard to find. The exact rationale for this is hard to determine, but it seems possible that it’s a topic that folks are generally uncomfortable discussing in public on account of potential future liability and compliance issues. In our exploration, we found two standardized dimensions (role-based access controls, audit logs), and one highly divergent dimension (resource-specific access controls): Role-based access controls (RBAC) are a highly standardized approach at this point. The core premise is that users are mapped to one or more roles, and each role is granted a certain set of permissions. For example, a role representing the customer support agent might be granted permission to deactivate an account, whereas a role representing the sales engineer might be able to configure a new account. Audit logs are similarly standardized. All access and mutation of resources should be tied in a durable log to the human who performed the action. These logs should be accumulated in a centralized, queryable solution. One of the core challenges is determining how to utilize these logs proactively to detect issues rather than reactively when an issue has already been flagged. Resource-level access controls are significantly less standardized than RBAC or audit logs. We found three distinct patterns adopted by companies, with little consistency across companies on which is adopted. Those three patterns for resource-level access control were: 3rd-party enrichment where access to resources is managed in a 3rd-party system such as Zendesk. This requires enriching objects within those systems with data and metadata from the product(s) where those objects live. It also requires implementing actions on the platform, such as archiving or configuration, allowing them to live entirely in that platform’s permission structure. The downside of this approach is tight coupling with the platform vendor, any limitations inherent to that platform, and the overhead of maintaining engineering teams familiar with both your internal technology stack and the platform vendor’s technology stack. 1st-party tool implementation where all activity, including creation and management of user issues, is managed within the core product itself. This pattern is most common in earlier stage companies or companies whose customer support leadership “grew up” within the organization without much exposure to the approach taken by peer companies. The advantage of this approach is that there is a single, tightly integrated and infinitely extensible platform for managing interactions. The downside is that you have to build and maintain all of that work internally rather than pushing it to a vendor that ought to be able to invest more heavily into their tooling. Hybrid solutions where a 3rd-party platform is used for most actions, and is further used to permit resource-level access within the 1st-party system. For example, you might be able to access a user’s data only while there is an open ticket created by that user, and assigned to you, in the 3rd-party platform. The advantage of this approach is that it allows supporting complex workflows that don’t fit within the platform’s limitations, and allows you to avoid complex coupling between your product and the vendor platform. Generally, our experience is that all companies implement RBAC, audit logs, and one of the resource-level access control mechanisms. Most companies pursue either 3rd-party enrichment with a sizable, long-standing team owning the platform implementation, or rely on a hybrid solution where they are able to avoid a long-standing dedicated team by lumping that work into existing teams.

Entering 2025, I decided to spend some time exploring the topic of agents. I started reading Anthropic’s Building effective agents, followed by Chip Huyen’s AI Engineering. I kicked off a major workstream at work on using agents, and I also decided to do a personal experiment of sorts. This is a general commentary on building that project. What I wanted to build was a simple chat interface where I could write prompts, select models, and have the model use tools as appropriate. My side goal was to build this using Cursor and generally avoid writing code directly as much as possible, but I found that generally slower than writing code in emacs while relying on 4o-mini to provide working examples to pull from. Similarly, while I initially envisioned building this in fullstack TypeScript via Cursor, I ultimately bailed into a stack that I’m more comfortable, and ended up using Python3, FastAPI, PostgreSQL, and SQLAlchemy with the async psycopg3 driver. It’s been a… while… since I started a brand new Python project, and used this project as an opportunity to get comfortable with Python3’s async/await mechanisms along with Python3’s typing along with mypy. Finally, I also wanted to experiment with Tailwind, and ended up using TailwindUI’s components to build the site. The working version supports everything I wanted: creating chats with models, and allowing those models to use function calling to use tools that I provide. The models are allowed to call any number of tools in pursuit of the problem they are solving. The tool usage is the most interesting part here for sure. The simplest tool I created was a get_temperature tool that provided a fake temperature for your location. This allowed me to ask questions like “What should I wear tomorrow in San Francisco, CA?” and get a useful respond. The code to add this function to my project was pretty straightforward, just three lines of Python and 25 lines of metadata to pass to the OpenAI API. def tool_get_current_weather(location: str|None=None, format: str|None=None) -> str: "Simple proof of concept tool." temp = random.randint(40, 90) if format == 'fahrenheit' else random.randint(10, 25) return f"It's going to be {temp} degrees {format} tomorrow." FUNCTION_REGISTRY['get_current_weather'] = tool_get_current_weather TOOL_USAGE_REGISTRY['get_current_weather'] = { "type": "function", "function": { "name": "get_current_weather", "description": "Get the current weather", "parameters": { "type": "object", "properties": { "location": { "type": "string", "description": "The city and state, e.g. San Francisco, CA", }, "format": { "type": "string", "enum": ["celsius", "fahrenheit"], "description": "The temperature unit to use. Infer this from the users location.", }, }, "required": ["location", "format"], }, } } After getting this tool, the next tool I added was a simple URL retriever tool, which allowed the agent to grab a URL and use the content of that URL in its prompt. The implementation for this tool was similarly quite simple. def tool_get_url(url: str|None=None) -> str: if url is None: return '' url = str(url) response = requests.get(url) soup = BeautifulSoup(response.content, 'html.parser') content = soup.find('main') or soup.find('article') or soup.body if not content: return str(response.content) markdown = markdownify(str(content), heading_style="ATX").strip() return str(markdown) FUNCTION_REGISTRY['get_url'] = tool_get_url TOOL_USAGE_REGISTRY['get_url'] = { "type": "function", "function": { "name": "get_url", "description": "Retrieve the contents of a website via its URL.", "parameters": { "type": "object", "properties": { "url": { "type": "string", "description": "The complete URL, including protocol to retrieve. For example: \"https://lethain.com\"", } }, "required": ["url"], }, } } What’s pretty amazing is how much power you can add to your agent by adding such a trivial tool as retrieving a URL. You can similarly imagine adding tools for retrieving and commenting on Github pull requests and so, which could allow a very simple agent tool like this to become quite useful. Working on this project gave me a moderately compelling view of a near-term future where most engineers have simple application like this running that they can pipe events into from various systems (email, text, Github pull requests, calendars, etc), create triggers that map events to templates that feed into prompts, and execute those prompts with tool-aware agents. Combine that with ability for other agents to register themselves with you and expose the tools that they have access to (e.g. schedule an event with tool’s owner), and a bunch of interesting things become very accessible with a very modest amount of effort: You could schedule events between two busy people’s calendars, as if both of them had an assistant managing their calendar Reply to your own pull requests with new blog posts, providing feedback on typos and grammatical issues Crawl websites you care about and identify posts you might be interested in Ask the model to generate a system model using lethain:systems, run that model, then chart the responses Add a “planning tool” which allows the model to generate a plan to guide subsequent steps in a complex task. (e.g. getting my calendar, getting a friend’s calendar, suggesting a time we could meet) None of these are exactly lifesaving, but each is somewhat useful, and I imagine there are many more fairly obvious ideas that become easy once you have the necessary scaffolding to make this sort of thing easy. Altogether, I think that I am convinced at this points that agents, using current foundational models, are going to create a number of very interesting experiences that improve our day to day lives in small ways that are, in aggregate, pretty transformational. I’m less convinced that this is the way all software should work going forward though, but more thoughts on that over time. (A bunch of fun experiments happening at work, but early days on those.)

While I frequently hear engineers bemoan a missing strategy, they rarely complete the thought by articulating why the missing strategy matters. Instead, it serves as more of a truism: the economy used to be better, children used to respect their parents, and engineering organizations used to have an engineering strategy. This chapter starts by exploring something I believe quite strongly: there’s always an engineering strategy, even if there’s nothing written down. From there, we’ll discuss why strategy, especially written strategy, is such a valuable opportunity for organizations that take it seriously. We’ll dig into: Why there’s always a strategy, even when people say there isn’t How strategies have been impactful across my career How inappropriate strategies create significant organizational pain without much compensating impact How written strategy drives organizational learning The costs of not writing strategy down How strategy supports personal learning and developing, even in cases where you’re not empowered to “do strategy” yourself By this chapter’s end, hopefully you will agree with me that strategy is an undertaking worth investing your–and your organization’s–time in. This is an exploratory, draft chapter for a book on engineering strategy that I’m brainstorming in #eng-strategy-book. As such, some of the links go to other draft chapters, both published drafts and very early, unpublished drafts. There’s always a strategy I’ve never worked somewhere where people didn’t claim there as no strategy. In many of those companies, they’d say there was no engineering strategy. Once I became an executive and was able to document and distribute an engineering strategy, accusations of missing strategy didn’t go away, they just shfited to focus on a missing product or company strategy. This even happened at companies that definitively had engineering strategies like Stripe in 2016 which had numerous pillars to a clear engineering strategy such as: Maintain backwards API compatibilty, at almost any cost (e.g. force an upgrade from TLS 1.2 to TLS 1.3 to retain PCI compliance, but don’t force upgrades from the /v1/charges endpoint to the /v1/payment_intents endpoint) Work in Ruby in a monorepo, unless it’s the PCI environment, data processing, or data science work Engineers are fully responsible for the usability of their work, even when there are product or engineering managers involved Working there it was generally clear what the company’s engineering strategy was on any given topic. That said, it sometimes required asking around, and over time certain decisions became sufficiently contentious that it became hard to definitively answer what the strategy was. For example, the adoptino of Ruby versus Java became contentious enough that I distributed a strategy attempting to mediate the disagreement, Magnitudes of exploration, although it wasn’t a particularly successful effort (for reasons that are obvious in hindsight, particularly the lack of any enforcement mechanism). In the same sense that William Gibson said “The future is already here – it’s just not very evenly distributed,” there is always a strategy embedded into an organization’s decisions, although in many organizations that strategy is only visible to a small group, and may be quickly forgotten. If you ever find yourself thinking that a strategy doesn’t exist, I’d encourage you to instead ask yourself where the strategy lives if you can’t find it. Once you do find it, you may also find that the strategy is quite ineffective, but I’ve simply never found that it doesn’t exist. Strategy is impactful In “We are a product engineering company!”, we discuss Calm’s engineering strategy to address pervasive friction within the engineering team. The core of that strategy is clarifying how Calm makes major technology decisions, along with documenting the motivating goal steering those decisions: maximizing time and energy spent on creating their product. That strategy reduced friction by eliminating the cause of ongoing debate. It was successful in resetting the team’s focus. It also caused several engineers to leave the company, because it was incompatible with their priorities. It’s easy to view that as a downside, but I don’t think it was. A clear, documented strategy made it clear to everyone involved what sort of game we were playing, the rules for that game, and for the first time let them accurately decide if they wanted to be part of that game with the wider team. Creating alignment is one of the ways that strategy makes an impact, but it’s certainly not the only way. Some of the ways that strategies support the creating organization are: Concentrating company investment into a smaller space. For example, deciding not to decompose a monolith allows you to invest the majority of your tooling efforts on one language, one test suite, and one deployment mechanism. Many interesting properties only available through universal adoption. For example, moving to an “N-1 policy” on backfilled roles is a significant opportunity for managing costs, but only works if consistently adopted. As another example, many strategies for disaster recovery or multi-region are only viable if all infrastructure has a common configuration mechanism. Focus execution on what truly matters. For example, Uber’s service migration strategy allowed a four engineer team to migrate a thousand services operated by two thousand engineers to a new provisioning and orchestration platform in less than a year. This was an extraordinarily difficult project, and was only possible because of clear thinking. Creating a knowledge repository of how your organization thinks. Onboarding new hires, particularly senior new hires, is much more effective with documented strategy. For example, most industry professionals today have a strongly held opinion on how to adopt large language models. New hires will have a strong opinion as well, but they’re unlikely to share your organization’s opinion unless there’s a clear document they can read to understand it. There are some things that a strategy, even a cleverly written one, cannot do. However, it’s always been my experience that developing a strategy creates progress, even if the progress is understanding the inherent disagreement preventing agreement. Inappropriate strategy is especially impactful While good strategy can accomplish many things, it sometimes feels that inappropriate strategy is far more impactful. Of course, impactful in all the wrong ways. Digg V4 remains the worst considered strategy I’ve personally participated in. It was a complete rewrite of the Digg V3.5 codebase from a PHP monolith to a PHP frontend and backend of a dozen Python services. It also moved the database from sharded MySQL to an early version of Cassandra. Perhaps worst, it replaced the nuanced algorithms developed over a decade with a hack implemented a few days before launch. Although it’s likely Digg would have struggled to become profitable due to its reliance on search engine optimization for traffic, and Google’s frequently changing search algorithm of that era, the engineering strategy ensured we died fast rather than having an opportunity to dig our way out. Importantly, it’s not just Digg. Almost every engineering organization you drill into will have it’s share of unused platform projects that captured decades of engineering years to the detriment of an important opportunity. A shocking number of senior leaders join new companies and initiate a grand migration that attempts to entirely rewrite the architecture, switch programming languages, or otherwise shift their new organization to resemble a prior organization where they understood things better. Inappropriate versus bad When I first wrote this section, I just labeled this sort of strategy as “bad.” The challenge with that term is that the same strategy might well be very effective in a different set of circumstances. For example, if Digg had been a three person company with no revenue, rewriting from scratch could have the right decision! As a result, I’ve tried to prefer the term “inappropriate” rather than “bad” to avoid getting caught up on whether a given approach might work in other circumstances. Every approach undoubtedly works in some organization. Written strategy drives organizational learning When I joined Carta, I noticed we had an inconsistent approach to a number of important problems. Teams had distinct standard kits for how they approached new projects. Adoption of existing internal platforms was inconsistent, as was decision making around funding new internal platforms. There was widespread agreement that we were decomposing our monolith, but no agreement on how we were doing it. Coming into such a permissive strategy environment, with strong, differing perspectives on the ideal path forward, one of my first projects was writing down an explicit engineering strategy along with our newly formed Navigators team, itself a part of our new engineering strategy. Navigators at Carta As discussed in Navigators, we developed a program at Carta to have explicitly named individual contributor, technical leaders to represent key parts of the engineering organization. This representative leadership group made it possible to iterate on strategy with a small team of about ten engineers that represented the entire organization, rather than take on the impossible task of negotiating with 400 engineers directly. This written strategy made it possible to explicitly describe the problems we saw, and how we wanted to navigate those problems. Further, it was an artifact that we were able to iterate on in a small group, but then share widely for feedback from teams we might have missed. After initial publishing, we shared it widely and talked about it frequently in engineering all-hands meetings. Then we came back to it each year, or when things stopped making much sense, and revised it. As an example, our initial strategy didn’t talk about artificial intelligence at all. A few months later, we extended it to mention a very conservative approach to using Large Language Models. Most recently, we’ve revised the artificial intelligence portion again, as we dive deeply into agentic workflows. A lot of people have disagreed with parts of the strategy, which is great: that’s one of the key benefits of a written strategy, it’s possible to precisely disagree. From that disagreement, we’ve been able to evolve our strategy. Sometimes because there’s new information like the current rapidly evolution of artificial intelligence pratices, and other times because our initial approach could be improved like in how we gated membership of the initial Navigators team. New hires are able to disagree too, and do it from an informed place rather than coming across as attached to their prior company’s practices. In particular, they’re able to understand the historical thinking that motivated our decisions, even when that context is no longer obvious. At the time we paused decomposition of our monolith, there was significant friction in service provisioning, but that’s far less true today, which makes the decision seem a bit arbitrary. Only the written document can consistently communicate that context across a growing, shifting, and changing organization. With oral history, what you believe is highly dependent on who you talk with, which shapes your view of history and the present. With writen history, it’s far more possible to agree at scale, which is the prerequisite to growing at scale rather than isolating growth to small pockets of senior leadership. The cost of implicit strategy We just finished talking about written strategy, and this book spends a lot of time on this topic, including a chapter on how to structure strategies to maximize readability. It’s not just because of the positives created by written strategy, but also because of the damage unwritten strategy creates. Vulnerable to misinterpretation. Information flow in verbal organizations depends on an individual being in a given room for a decision, and then accurately repeating that information to the others who need it. However, it’s common to see those individuals fail to repeat that information elsewhere. Sometimes their interpretation is also faulty to some degree. Both of these create significant problems in operating strategy. Two-headed organizations Some years ago, I started moving towards a model where most engineering organizations I worked with have two leaders: one who’s a manager, and another who is a senior engineer. This was partially to ensure engineering context was included in senior decision making, but it was also to reduce communication errors. Errors in point-to-point communication are so prevalent when done one-to-one, that the only solution I could find for folks who weren’t reading-oriented communicators was ensuring I had communicated strategy (and other updates) to at least two people. Inconsistency across teams. At one company I worked in, promotions to Staff-plus role happened at a much higher rate in the infrastructure engineering organization than the product engineering team. This created a constant drain out of product engineering to work on infrastructure shaped problems, even if those problems weren’t particularly valuable to the business. New leaders had no idea this informal policy existed, and they would routinely run into trouble in calibration discussions. They also weren’t aware they needed to go argue for a better policy. Worse, no one was sure if this was a real policy or not, so it was ultimately random whether this perspective was represented for any given promotion: sometimes good promotions would be blocked, sometimes borderline cases would be approved. Inconsistency over time. Implementing a new policy tends to be a mix of persistent and one-time actions. For example, let’s say you wanted to standardize all HTTP operations to use the same library across your codebase. You might add a linter check to reject known alternatives, and you’ll probably do a one-time pass across your codebase standardizing on that library. However, two years later there are another three random HTTP libraries in your codebase, creeping into the cracks surrounding your linting. If the policy is written down, and a few people read it, then there’s a number of ways this could be nonetheless prevented. If it’s not written down, it’s much less likely someone will remember, and much more likely they won’t remember the rationale well enough to argue about it. Hazard to new leadership. When a new Staff-plus engineer or executive joins a company, it’s common to blame them for failing to understand the existing context behind decisions. That’s fair: a big part of senior leadership is uncovering and understanding context. It’s also unfair: explicit documentation of prior thinking would have made this much easier for them. Every particularly bad new-leader onboarding that I’ve seen has involved a new leader coming into an unfilled role, that the new leader’s manager didn’t know how to do. In those cases, success is entirely dependent on that new leader’s ability and interest in learning. In most ways, the practice of documenting strategy has a lot in common with succession planning, where the full benefits accrue to the organization rather than to the individual doing it. It’s possible to maintain things when the original authors are present, appreciating the value requires stepping outside yourself for a moment to value things that will matter most to the organization when you’re no longer a member. Information herd immunity A frequent objection to written strategy is that no one reads anything. There’s some truth to this: it’s extremely hard to get everyone in an organization to know something. However, I’ve never found that goal to be particularly important. My view of information dispersal in an organization is the same as Herd immunity: you don’t need everyone to know something, just to have enough people who know something that confusion doesn’t propagate too far. So, it may be impossible for all engineers to know strategy details, but you certainly can have every Staff-plus engineer and engineering manager know those details. Strategy supports personal learning While I believe that the largest benefits of strategy accrue to the organization, rather than the individual creating it, I also believe that strategy is an underrated avenue for self-development. The ways that I’ve seen strategy support personal development are: Creating strategy builds self-awareness. Starting with a concrete example, I’ve worked with several engineers who viewed themselves as extremely senior, but frequently demanded that projects were implemented using new programming languages or technologies because they personally wanted to learn about the technology. Their internal strategy was clear–they wanted to work on something fun–but following the steps to build an engineering strategy would have created a strategy that even they agreed didn’t make sense. Strategy supports situational awareness in new environments. Wardley mapping talks a lot about situational awareness as a prerequisite to good strategy. This is ensuring you understand the realities of your circumstances, which is the most destructive failure of new senior engineering leaders. By explicitly stating the diagnosis where the strategy applied, it makes it easier for you to debug why reusing a prior strategy in a new team or company might not work. Strategy as your personal archive. Just as documented strategy is institutional memory, it also serves as personal memory to understand the impact of your prior approaches. Each of us is an archivist of our prior work, pulling out the most valuable pieces to address the problem at hand. Over a long career, memory fades–and motivated reasoning creeps in–but explicit documentation doesn’t. Indeed, part of the reason I started working on this book now rather than later is that I realized I was starting to forget the details of the strategy work I did earlier in my career. If I wanted to preserve the wisdom of that era, and ensure I didn’t have to relearn the same lessons in the future, I had to write it now. Summary We’ve covered why strategy can be a valuable learning mechanism for both your engineering organization and for you. We’ve shown how strategies have helped organizations deal with service migrations, monolith decomposition, and right-sizing backfilling. We’ve also discussed how inappropriate strategy contributed to Digg’s demise. However, if I had to pick two things to emphasize as this chapter ends, it wouldn’t be any of those things. Rather, it would be two themes that I find are the most frequently ignored: There’s always a strategy, even if it isn’t written down. The single biggest act you can take to further strategy in your organization is to write down strategy so it can be debated, agreed upon, and explicitly evolved. Discussions around topics like strategy often get caught up in high prestige activities like making controversial decisions, but the most effective strategists I’ve seen make more progress by actually performing the basics: writing things down, exploring widely to see how other companies solve the same problem, accepting feedback into their draft from folks who disagree with them. Strategy is useful, and doing strategy can be simple, too.

In my career, the majority of the strategy work I’ve done has been in non-executive roles, things like Uber’s service migration. Joining Calm was my first executive role, where I was able to not just propose, but also mandate, strategy. Like almost all startups, the engineering team was scattered when I joined. Was our most important work creating more scalable infrastructure? Was our greatest risk the failure to adopt leading programming languages? How did we rescue the stuck service decomposition initiative? This strategy is where the engineering team and I aligned after numerous rounds of iteration, debate, and inevitably some disagreement. As a strategy, it’s both basic and also unambiguous about what we valued, and I believe it’s a reasonably good starting point for any low scalability-complexity consumer product. This is an exploratory, draft chapter for a book on engineering strategy that I’m brainstorming in #eng-strategy-book. As such, some of the links go to other draft chapters, both published drafts and very early, unpublished drafts. Reading this document To apply this strategy, start at the top with Policy. To understand the thinking behind this strategy, read sections in reverse order, starting with Explore, then Diagnose and so on. Relative to the default structure, this document has one tweak, folding the Operation section in with Policy. More detail on this structure in Making a readable Engineering Strategy document. Policy & Operation Our new policies, and the mechanisms to operate them are: We are a product engineering company. Users write in every day to tell us that our product has changed their lives for the better. Our technical infrastructure doesn’t get many user letters–and this is unlikely to change going forward as our infrastructure is relatively low-scale and low-complexity. Rather than attempting to change that, we want to devote the absolute maximum possible attention to product engineering. We exclusively adopt new technologies to create valuable product capabilities. We believe our technology stack as it exists today can solve the majority of our current and future product roadmaps. In the rare case where we adopt a new technology, we do so because a product capability is inherently impossible without adopting a new technology. We do not adopt new technologies for other reasons. For example, we would not adopt a new technology because someone is interested in learning about it. Nor would we adopt a technology because it is 30% better suited to a task. We write all code in the monolith. It has been ambiguous if new code (especially new application code) should be written in our JavaScript monolith, or if all new code must be written in a new service outside of the monolith. This is no longer ambiguous: all new code must be written in the monolith. In the rare case that there is a functional requirement that makes writing in the monolith implausible, then you should seek an exception as described below. Exceptions are granted by the CTO, and must be in writing. The above policies are deliberately restrictive. Sometimes they may be wrong, and we will make exceptions to them. However, each exception should be deliberate and grounded in concrete problems we are aligned both on solving and how we solve them. If we all scatter towards our preferred solution, then we’ll create negative leverage for Calm rather than serving as the engine that advances our product. All exceptions must be written. If they are not written, then you should operate as if it has not been granted. Our goal is to avoid ambiguity around whether an exception has, or has not, been approved. If there’s no written record that the CTO approved it, then it’s not approved. Proving the point about exceptions, there are two confirmed exceptions to the above strategy: We are incrementally migrating to TypeScript. We have found that static typing can prevent a number of our user-facing bugs. TypeScript provides a clean, incremental migration path for our JavaScript codebase, and we aim to migrate the entirety over the next six months. Our Web engineering team is leading this migration. We are evaluating Postgres Aurora as our primary database. Many of our recent production incidents are caused by index scans for tables with high write velocity such as tracking customer logins. We believe Aurora will perform better under these workloads. Our Infrastructure engineering team is leading this initiative. Diagnose The current state of our engineering organization: Our product is not limited by missing infrastructure capabilities. Reviewing our roadmap, there’s nothing that we are trying to build today or over the next year that is constrained by our technical infrastructure. Our uptime, stability and latency are OK but not great. We have semi-frequent stability and latency issues in our application, all of which are caused by one of two issues. First, deploying new code with a missing index because it performed well enough in a test environment. Second, writes to a small number of extremely large, skinny tables have become expensive in combination with scans over those tables’ indexes. Our infrastructure team is split between supporting monolith and service workflows. One way to measure technical debt is to understand how much time the team is spending propping up the current infrastructure. Today, that is meaningful but not overwhelming work for our team of three infrastructure engineers supporting 30 product engineers. However, we are finding infrastructure engineers increasingly pulled into debugging incidents for components moved out of the central monolith into our service architecture. This is partially due to increased inherent complexity, but it’s more due to exposing lack of monitoring and ambiguous accountability in services’ production incidents. Our product and executive stakeholders experience us as competing factions. Engineering exists to build and operate software in the company. Part of that is being easy to work with. We should not necessarily support every ask from Product if we believe they are misaligned with Engineering’s goals (e.g. maintaining security), but it should generally provide a consistent perspective across our team. Today, our stakeholders believe they will get radically different answers to basic questions of capabilities and approach depending on who they ask. If they try to get a group of engineers to agree on an approach, they often find we derail into debate about approach rather than articulating a clear point of view that allows the conversation to move forward. We’re arguing a particularly large amount about adopting new technologies and rewrites. Most of our disagreements stem around adopting new technologies or rewriting existing components into new technology stacks. For example, can we extend this feature or do we have to migrate it to a service before extending it? Can we add this to our database or should we move it into a new Redis cache instead? Is JavaScript a sufficient programming language, or do we need to rewrite this functionality in Go? This is particularly relevant to next steps around the ongoing services migration, which has been in-flight for over a year, but is yet to move any core production code. We are spending more time on infrastructure and platform work than product work. This is the combination of all the above issues, from the stability issues we are encountering in our database design, to the lack of engineering alignment on execution. This places us at odds with stakeholder expectation that we are predominantly focused on new product development. Explore Calm is a mobile application that guides users to build and maintain either a meditation or sleep habit. Recommendations and guidance across content is individual to the user, but the content is shared across all customers and is amenable to caching on a content delivery network (CDN). As long as the CDN is available, the mobile application can operate despite inability to access servers (e.g. the application remains usable from a user’s perspective, even if the non-CDN production infrastructure is unreachable). In 2010, enabling a product of this complexity would have required significant bespoke infrastructure, along with likely maintaining a physical presence in a series of datacenters to run your software. In 2020, comparable applications are generally moving towards maintaining as little internal infrastructure as possible. This perspective is summarized effectively in Intercom’s Run Less Software and Dan McKinley’s Choose Boring Technology. New companies founded in this space view essentially all infrastructure as a commodity bought off your cloud provider. This even extends to areas of innovation, such as machine learning, where the training infrastructure is typically run on an offering like AWS Bedrock, and the model infrastructure is provided by Anthropic or OpenAI.