The Future of Engineering Teams in the Age of AI
Written on
Chapter 1: The Changing Landscape of Engineering Teams
As we delve into the effects of AI coding tools on engineering roles, a critical question arises: if these tools lessen the need for engineers, how should we allocate our budgets?
This section is part of a six-part series examining the potential transformations within product engineering teams due to generative AI tools aimed at developers, such as GitHub Copilot, ChatGPT, and Amazon CodeWhisperer. In the previous part, we discussed various aspects, including:
- The possibility of altering the traditional 5:1 engineer-to-product-manager ratio.
- The influence of tools like GitHub Copilot and AWS Amplify Studio on product development, encouraging engineers to pivot towards design, architecture, and integration rather than just coding.
- The support AI tools can provide teams grappling with outdated technology, simplifying complex porting and refactoring processes.
- The potential for AI tools to unify mobile and web app development, reducing redundancy and addressing skill gaps across platforms.
- The ramifications of coding automation on junior developers and career progression in engineering.
The emergence of large language models (LLMs) has undeniably shifted the paradigm, presenting significant opportunities for generative AI tools to enhance our work. Yet, as McKinsey pointed out, there will be both beneficiaries and those left behind in this transformative era.
Despite optimistic trends in the U.S. economy and a remarkable rebound in profitability from major tech firms like Microsoft, Google, and Meta, global economic uncertainty—driven by inflation—will likely influence how companies respond to these new opportunities. In a more favorable economic climate, many businesses would view this landscape as a chance for growth, propelled by the exuberance of recent investment trends.
However, the current reality paints a different picture. Venture capital funding saw a staggering 53% decline in Q1 2023 compared to the previous year. The fintech sector, in particular, faced a troubling 48% drop in funding for Q2 2023, the lowest level since 2017. The era of easy capital is over—unless you're a genuine deep-tech AI company, which currently seems to be in a frenzy of investment.
Given the tougher economic conditions, many startups have been forced to trim discretionary spending to survive. For numerous companies, generative AI coding tools may now be viewed as a vital strategy to conserve cash and extend their operational runway.
For organizations in a more stable financial position, this opportunity could allow them to either maintain or expand their product engineering budgets while restructuring their teams to adapt to the evolving landscape.
Video Description: In this video, we explore whether AI can code a data engineering project using ChatGPT to develop a Python application.
Chapter 2: Exploring Different Scenarios
Let’s engage in a thought experiment and consider how adopting a 1:1 engineer-to-product-manager ratio might play out across three distinct scenarios:
- A financially robust company eager to grow without losing any talented engineers.
- A company facing financial constraints that opts to maintain its current output while significantly reducing its engineering workforce.
- A company that seeks to keep its budget steady while transforming its team structure to align with the new ratios, thereby increasing output without additional costs.
These models are, of course, oversimplified and serve merely as caricatures to spur thoughtful discussion.
Starting Scenario
Imagine a product and engineering team composed of 100 individuals. Based on our previous engineer-to-product-manager ratio, we assume each team comprises one product manager and five engineers. Additionally, there are two 'other' roles in each team (such as data engineers, QA, design, UX, devops, etc.), leading to a total of 5 to 9 members per team.
To support these teams, we also account for one central or management role per team, necessary to prevent organizational overload and facilitate communication.
Thus, within our 100-person product engineering team (adhering to the 5:1 ratio), we have 89 team members spread across 11 teams, with 11 product managers and 55 engineers.
Scenario 1: Growth
Let’s envision an optimistic scenario where our first company is exceptionally profitable and committed to retaining all its engineers. Adopting the new 1:1 ratio creates an exciting opportunity to expand from 11 productive teams to 55, potentially allowing the organization to fulfill its entire roadmap for the year.
However, this unbounded optimism presents challenges, as the company would need to hire an additional 44 product managers, significantly impacting other roles like QA, UX, and design due to the increased workload. Consequently, the organization would balloon to nearly three times its original size, with a corresponding salary increase from £10 million to £27 million. Yet, they would retain the same number of engineers.
Scenario 2: Save Cash, Cut Costs
In contrast, companies grappling with financial challenges often ask, "Where can we cut costs?" Tech leaders will need to determine how to keep their teams small while still achieving the desired outcomes.
With our new ratios, this inquiry yields a stark conclusion. The enhanced productivity per developer suggests that we can drastically reduce the number of engineers, preserving the existing 11 teams and their commitments while slicing 44 engineering roles from the budget.
While our basic ratios imply that other team and management roles remain unchanged, the reality is likely more complex. The overall headcount would drop from 100 to 55, yielding benefits such as improved communication, reduced complexity, and enhanced efficiency. Surprisingly, the smaller team might outperform its larger counterpart.
Scenario 3: Zero Sum: Maintain Budget
The most plausible outcome might be a moderately successful company that avoids drastic financial decisions. Here, budgets remain consistent, providing an opportunity to increase outcomes without raising costs.
In this scenario, product and tech leaders recognize that they can maintain a workforce of 100 employees while expanding from 11 to 20 teams. This growth necessitates hiring nine new product roles and boosting other roles from 22 to 40. However, this expansion must occur without increasing costs, resulting in a cut of 35 engineering positions due to enhanced productivity.
While this shift may be disruptive, it is significantly less severe than the previous scenarios. It preserves the overall team size and budget but alters the composition.
At its core, our exploration centers around the question: "Will generative AI tools reduce the number of developers or increase the number of product managers?" It's essential to remember that the 1:1 ratio is hypothetical, likely skewing towards a more realistic 3:1 or 2:1 based on operational realities and productivity research.
In Part 5, we will delve into:
- The effects of coding automation tools on businesses of varying sizes, from startups to large corporations.
- The challenges faced by outsourced development companies in light of transformative AI tools as hiring dynamics, costs, and value propositions evolve.
- The irreplaceable role of human engineers in the AI era.
Stay tuned for Part 5, and don't forget to check out the other articles in this series: Part 1, Part 2, Part 3.
P.S. If you're enjoying these insights on team dynamics, listen to my Teamcraft podcast, where my co-host Andrew Maclaren and I discuss what makes teams thrive.
Video Description: Discover how AI can be utilized in structural engineering, exploring its potential applications and benefits in this informative discussion.