Through 2018, 10X has been slowly growing its team. We're now five full-time members with a few offer letters out. Compared to a year ago, that's triple the headcount — but by startup standards, it's a slow build, I think.
"How do you build a team capable of inventing things people want, and the leverage to deliver them?" The intention to create a culture of fierce product focus has been consistent from the beginning.
The key forces in inventing "things" and "traffic" are, I believe, the power to create and the power to engineer. I loosely define this as "the power of creators" (I consider engineering inherently creative). And the role of management in the team we're building distills down to one thing: "maximize the throughput produced by creators working as a team."
Throughput per unit time is also called speed. That speed matters is obvious — and in hiring and team design alike, I hold myself to the question: "Can this maximize throughput?"
As we grow the team going forward, the top priority is clearly "drawing in people who resonate with our culture and the future we're trying to build." Anyone whose skills and timing align, I'd love to have join us (reach out here).
At the same time: how do we design the systems that help members who join create real chemistry, respond to shifting direction, and maximize throughput? This is the question occupying my mind lately. Let me think through it.
The WHY of evaluation
Every team has an "evaluation system," but why is evaluation necessary in the first place? For a small team like ours, I see two reasons:
- Feedback: Create opportunities for individuals to review their own work and behavior — from themselves and from others — and apply feedback in the direction of "maximizing throughput."
- Compensation: Within the relationship between company and individual, find a "mutually satisfying compensation level" so both sides can focus on throughput.
I'll work through these two purposes separately.
1 / Feedback
My conclusion: for the goal of maximizing throughput, "the finer the feedback, the better" and "feedback discussions should be kept separate from compensation discussions."
Take the "quarterly review" used in many large companies — the intervals are far too long. Reconstructing what you were thinking, what you did, and what the results were three months ago is costly for everyone.
What does the state of "maximum throughput" actually require? Something like this:
- Plans receive feedback in the moment and can be refined into better plans
- Insights during execution and progress updates receive feedback in the moment, enabling course corrections
- Results receive feedback so that budgets and strategies can be revised, and even large decisions like a pivot can be made
Under OKR or MBO systems where quarterly reviews are the norm, I have memories of being blindsided by feedback on actions I took three months ago, thinking "why now?" That feedback probably contained discussion relevant to compensation as well — but as feedback in service of throughput, it arrived far too late. The intent behind any given evaluation — feedback or compensation — should be made clear and separated.
At 10X, we run feedback through three main channels:
- On plans: Everything — business strategy, product specs, even meeting agendas — is drafted in text and shared for review (either requested from specific people or open to all). We also run a one-week product release cycle, and after each release, things we didn't get to and new discoveries are fed into the next sprint.
- On technical issues: About once a month, engineers hold a technical review to discuss structural technical debt and priorities that are hard to surface in regular pull request reviews.
- Casual conversation: Beyond scheduled feedback, we keep people working side by side so they can naturally voice what they notice. Casual conversation enables feedback at any granularity or layer of the organization. Casual conversation is a privilege of staying small.
Also: any discussion or feedback that isn't written down becomes siloed knowledge and adds nothing to throughput. I treat this as equivalent to "not existing" — which is why I play the role of the person who constantly nags everyone to "document it and share it."
2 / Compensation
This is far harder to design than feedback. There are multiple dilemmas to navigate in thinking about "what mutually satisfying compensation looks like," and each is extremely difficult on its own — and they're all tangled together. At 10X we're primarily working through the following three:
Cash flow vs. burn rate cap · Market value vs. performance value · What the individual needs vs. what the company can realistically pay · etc.
a. Cash flow vs. burn rate cap
Member compensation comes from the company's existing and generated resources — so as long as those resources are limited, the compensation ceiling can't be raised. Compensation has to be designed against that balance and a forward-looking financial plan. Fundraising provides resources to support this until cash flow is generated, but at some point generating significant free cash flow is a must. If you want "the best people, with appropriate compensation," then challenging a large market with genuine risk is the minimum bar for a company. Like the Kakin Kingdom and the Hunters heading for the Dark Continent with financial gain as their motivation.
(Note: Hunter x Hunter reference — in the manga, a wealthy kingdom and elite hunters embark on a dangerous frontier expedition, with the promise of enormous resources as motivation.)
Continuously growing resources — that is, sustainably increasing cash flow and raising the burn rate ceiling — comes down entirely to "market selection and execution within it." The question "Will you challenge a large market?" sets the cap on this tension. And this is not a question for the company-employee relationship — it is a question that resolves entirely within the founder. This is where "perspective" is demanded.
Once a founder commits to a big challenge, this tension can be translated one step further: can you design the financing and business milestones to reach large cash flow? At 10X, we've made our direction explicit: "challenge a large market, invent something 10x, build a big business." Rather than fine-grained evaluation, we want to grow the resource base and increase member compensation.
b. Market value vs. performance value
The approach of "estimate both market value and the performance value of working at your company, and design appropriate compensation from both angles" looks fair on the surface — but the difficulty lies in the fact that "estimating value" is practically impossible. Currently at 10X, we're not actively incorporating this perspective into compensation design.
Market value is expressed as an annual salary. It's relatively easy to understand roughly where you land in "the market = the aggregate of other companies' valuations." But that's only surface level. Real value should be calculated by evaluating and estimating "what throughput is generated by having this person here" — the performance value.
This performance value, outside of jobs like sales or trading where work can be expressed in monetary terms, is almost impossible to calculate. For example, a candidate who can command a salary of X yen on the job market carries two layers of uncertainty:
- "Whether they can perform consistently at our company" can only be known by working together
- "Whether that performance carries the same value to our company as it does in the market" is not guaranteed
What can't be measured can't be measured.
This is an anxiety point for both company and employee. And precisely because of that, accepting "previous salary" or "the offer on the market" without thinking through it often makes it harder for both sides to feel genuinely satisfied. 10X wants to keep hiring people who are happy with both our culture and their work, and who want to leave a real impact on the world. The current team comes to work because they want to deliver results (in practice, every member here is self-driven). Rather than forcing performance evaluations, we've taken a clear-cut stance: "for now, we won't do compensation-driven performance evaluation — let's reach for the unexplored together and earn a big reward." We're not currently designing compensation around market benchmarks or performance assessments. (This perspective will certainly shift significantly as the organization matures. I want to keep it as a note about current thinking only.)
c. What the individual needs vs. what the company can realistically pay
What the company can pay depends on its resources, as covered in (a). On the other hand, "what each individual needs" depends on where they live and their lifestyle, but more significantly — and less controllably — on their household situation. For example, whether someone has dependents significantly changes how much cash buffer they need to feel comfortable. As a father of two young children at home, I know this firsthand.
Even as a startup, we want to ensure "enough to live a standard life in Tokyo" as a minimum. But that alone isn't enough to maximize throughput. For example, young children often come down with fevers suddenly. Parents frequently need to take time off work to handle it. Allowing them to take paid leave without guilt is of course important — but could we also offer them the option to use a sick-childcare service (which costs several tens of thousands of yen per month and is hard for individuals to afford) to "buy more time for throughput"? That seems like the better path.
How to balance household and work is a personal choice we respect. But "having constraints due to family circumstances, with no means to address them" is a throughput limiter — and building the tools to remove those constraints into compensation design is something I want to pursue. (This is still in the planning stage, but I want to turn it into a concrete initiative soon.)
10X's current state and direction
Through the above thinking, our compensation system currently operates as follows:
- Salaries are determined based on business and company progress, cash flow, and financial plans. They are revised as the stage changes.
- Currently, all employees receive a uniform salary. We don't publish the amount out of privacy considerations, but we share the current figure and future projections at the first interview.
- Stock options and equity are designed as long-term incentives. Details are also shared at the first interview and updated as appropriate.
- Evaluation and compensation design should evolve like a living thing in response to the state of the organization — it is, in fact, exactly what I think of as a product, and a subject I love.
Dreams and vision matter, but they're perishable — they can't sustain people on their own. Having worked at an NPO — literally a vessel full of dreams — I know this in my bones as a formative experience.
Realistic, mutually satisfying evaluation and compensation design is a must for maximizing throughput over the long run. I want to keep turning these gears in parallel with the product.
