Resources  /  Blog  /  Arkestro Management Lessons
Thought Leadership

Arkestro Management Lessons

November 12, 2019

Contributor Post from Edmund Zagorin

In building and scaling Arkestro, our team has worked hard to create an intentional culture where people are able to set individual and team goals that align with business metrics. We aim to build a company where anyone with vision and motivation who shares our values can do their best work, where we celebrate excellence and reward impact.

After reading the leadership books by John Doerr, Ben Horowitz and Andy Grove, I wanted to share a bit about how Arkestro approaches setting business goals and measuring progress.

Growth is exciting but also challenging, (the SV cliche being “good problems to have”) and requires particular attention by managers to inter-personal dynamics and emotional intelligence. Even at an AI company, we have to be obsessed with our employee experience and take an attitude of iteration, candor, transparency and continuous improvement.

Reflecting on the past year of Arkestro, I wanted to take stock of some learnings that may be helpful to other leadership teams in startups and large companies. Candidly, I also wanted to share some of the things we got wrong, how we recognized what wasn’t working, and how we learned and improved.

Values & Mission – Best Practices

      • Values and mission only truly matter if they impact operational decisions on a weekly basis

      • Mission and values should be reviewed as part of every team summit and Board meeting

      • Mission and values should be reviewed/updated with a team on an annual basis to ensure that they remain accurate and reflective of our operational decision-making

      • Mission and values should be concise, memorable and as specific as possible without being limiting

What we got wrong & learned

When we started, we had 7 values and a paragraph-length mission statement. Now our core values are: thoughtful, proactive, collaborative, and visionary. Our mission statement is: To align supplier negotiation processes with operational excellence. If our platform helps our customers align with their supplier, then we are doing our job correctly. Our customers’ primary pain point is executing a complex multi-dimensional buying and partnership decision, hence “thoughtful”. Our guiding design principle is that the most desired outcome should involve fewer steps by always having the application make a suggestion, or a prompt, hence “proactive”. We are fundamentally a product for teams to use, hence “collaborative”. We drive the most value by deploying at scale, hence “visionary”. Earlier in the company, this was “Promethean vision”, but there were persistent questions about what exactly the relevance of Prometheus was to a procurement platform, so we made it simpler.

Management culture – Best practices

      • All managers have weekly 1:1s with direct reports

      • The CEO has quarterly 1:1s with every member of the company

      • Standard 1:1 agenda is the following questions: How are you? What’s going well? What can improve? How can I help?

      • Quarterly targets measured at individual, team and company level, fully transparent

      • Use the Objectives & Key Result (OKRs) framework described in John Doerr’s book Measure What Matters

      • OKRs should be drafted by each team member and iterated with the manager before being presented to full team

      • Each employee is able to say: “If I am kicking ass, it’s because the following three metrics look like this… If my team is kicking ass, the following three metrics look like this… If the company is kicking ass, the following three metrics look like this…”

      • Good objectives are directional and aligned with business goals and good key results are both quantitative aspirational

      • Objectives should be compared against business OKRs, company mission and values

What we got wrong & learned from

Early on we did a lot of all-hands meetings to measure progress and discuss initiatives. These were inefficient and most participants were unable to express themselves due to a large number of participants. We quickly realized that people could not or did not express their best ideas in large group settings.

As a result, good ideas came out most frequently during ad hoc 1:1 discussions. We leaned into 1:1s and away from larger group meetings and all-hands. Making 1:1s a regular part of our weekly schedule has decreased the need for meetings and made our all-hands meetings way more effective. People also feel more heard when they can share their opinion 1:1, and smart ideas get visibility & implemented faster.

Meeting Culture – Best Practices

      • Meetings must start within 60 seconds of the calendared event

      • If someone schedules a meeting and is more than 60 seconds late without informing the other participants, the meeting is default canceled and must be rescheduled

      • No interrupting people. If an interruption has occurred speaker must indicate that there is more that must be heard. To prevent monologue, all speakers are encouraged to frequently pause for questions or comments

      • The default meeting length is 30 minutes. More than 3 people mean the meeting is 45 minutes. A best practice is to end meetings 5 minutes early

      • Meetings over an hour must have unique urgency and special justification

      • All presentation materials must be shared for more than 24 hours in advance

      • All meetings have agenda, owner/MC, notetaker and clear impact on business goals

What We got wrong & learned from

Adding people to conference calls/meetings/email threads for visibility purposes is an expensive way to share information. Much easier to have fewer meeting participants, share action items via shared doc/Slack then to ask anyone to be a passive participant in any meeting.

1:1s can go a long way to prevent a culture of passive participation, but having an agenda and an owner for meetings that start on time ensures that everyone who needs to engage has the opportunity to add value/improve project plans. Earlier on we frequently had meetings that start and end late, included many passive participants who each contributed (relatively) less than they wanted to. Now we have shorter meetings with fewer participants who each say more ideas.

Knowledge Culture – Best Practices

      • All repeatable processes must be documented in our internal wiki knowledge base, with clearly defined roles, steps and expected cycle time and expected milestone completion dates

      • All customer-facing processes must go through a rehearsal stage where we internally role play as the customer to work through the steps in the process

      • All customer-facing processes must go through an iteration phase where they are meaningfully improved

      • All processes must have a recurring calendared date when they are reviewed for continuous improvement

      • All new features must go through QA and user testing with a real professional negotiator

      • All customer process cycles must be connected to our analytics platforms and measured using a layer of automation that is connected to a mutually visible dashboard

      • All customer questions answers in text are added to shared internal FAQ

      • All questions received more than 3x become a Help Desk article


There is probably nothing more superficially boring than the design of a company’s meetings and knowledge practices, and yet they are absolutely vital to building a workplace where people can work without distraction, preserve and leverage their specialized focus, and propose ideas in a framework that can both hear them and implement them.

We’re by no means perfect and we have a long way to go, but this is something that we have focused on improving over the past year as we’ve grown our team. We wanted to share these insights in case they proved valuable to other early-stage startups that face the fundamental struggle: prioritize with clarity of purpose, iterate quickly, and execute on big goals with short timelines. If having shorter meetings makes the biggest difference to that ultimate goal, then it has suddenly become interesting to me.