Running a public benefit corporation, and integrating 25 acquisitions: Shanaz Hemmati of ZenBusiness

Most companies that talk about giving back do it through a values page. A public benefit corporation puts it in the charter, which changes what happens in the annual planning cycle.

Shanaz Hemmati is COO of ZenBusiness, a platform that handles company formation, legal, finance and go to market for small businesses. Before that she was an early employee at HomeAway, which she helped take public while acquiring more than 25 companies along the way, before HomeAway itself was acquired by Expedia.

She is also that rarer kind of operator: a COO with roots in engineering, infrastructure and data. She started at ZenBusiness as CIO before moving into the COO seat.

What a public benefit corporation actually commits you to

Michael opens on the corporate structure, and Hemmati's explanation is the practical version rather than the legal one.

A public benefit corporation is a type of corporation, most commonly established in Delaware, where the majority of companies incorporate their domestic entity. ZenBusiness is a corporation of that type. Her account of why is direct: the founders, particularly herself and CEO Ross Buhrdorf, wanted to know what they would give back, how they would support their community, and what they could do differently to have an impact.

The part operators should note is the enforcement mechanism. It is part of the charter, so it obligates them. They have to maintain activities and programs around it on an ongoing basis.

Which means it shows up in planning. Every year, as they plan the following one, there are initiatives specifically aligned to the public benefit commitment. Charitable donations and which organizations to support. Volunteering, which is complicated by having employees in 30 states, so they support it across many locations.

And then new programs. The one she describes is a customer grant scheme launched the previous year: applications come in monthly, they filter them, and depending on the annual allocation they select three or four recipients for a $5,000 grant to grow their business.

Michael's read on the underlying business is worth stating, because it explains why the grant program is consistent rather than decorative. ZenBusiness removes the back office, formation and ongoing compliance work so people can do the thing they are actually good at. Hemmati's own reason for joining was wanting to help people who have an idea and want to be their own boss.

From CIO to COO

Michael notes their portfolios are similar, both covering general and administrative functions alongside technical ones like infrastructure engineering and data science, and asks what drove the title change.

Her answer is a good description of how roles form in early companies. In year one they were mostly researching, testing and talking to potential customers before settling on being a platform. Her initial remit with Buhrdorf was the technology, the data, the internal tooling.

Once the strategy solidified, the division of labor got redrawn, and the redraw was based on evidence from that first year about what each of them was actually good at. With few people and limited resources, both had to be hands on regardless of title. The question became what each could do hands on and get done, versus where to find people to help.

The scope she ended up with is broad. The CTO reports to her, as do customer success, HR, legal, program management, procurement, and corporate citizenship. Her framing of managing senior leaders is one line: they have full responsibility for their areas, and she is there to enable them.

Buhrdorf, as CEO, has the CFO, CMO and chief product officer, and focuses on strategy, business development and corporate development. Finance reported to Hemmati until they hired a CFO the previous year, at which point it moved, which she attributes to the stage of the company rather than to anything about the function.

The phrase she borrows from a colleague is the one every COO will recognize: chief other officer. Things come up, there is nobody obvious to own them, so she takes ownership for a period and then decides whether it needs to move or whether it needs an owner hired for it.

Michael's observation is that this is also a decent method for working out what to hire next.

Build nothing until you have tested it

Michael asks about the wizard behind the curtain approach, doing things manually while presenting them as automated, so you do not overbuild.

Hemmati confirms they used it, and her reasoning has an unusually honest premise. Both she and Buhrdorf came from engineering, and in past lives at other companies they had built things nobody wanted.

So the discipline is: automate genuinely, because nobody should keep doing manual work a machine can do, but do not build something complex before testing it with actual customers. Research, feedback, then expansion. She says that is still how they operate.

Where AI helps operations, and where it does not

Her caution on AI is specific rather than generic, and it comes from the business model. ZenBusiness holds entity data for its customers, and the ease of the platform depends on that data being reused when a customer signs up for the next product. So how they use external AI tools has to be handled carefully.

Michael describes the state of the COO conversation at the time, which is that everyone started with a policy question, could not answer it, and settled on telling people not to put anything confidential in.

Hemmati's version adds the second half. Do not put anything confidential in, and go play with it. She was encouraging everyone in the company to try it on whatever task was in front of them, because prompting is a skill that requires practice, and because you learn things in the process that change what you were doing anyway.

Her framing of the value is the useful part for operators. Asked about challenges, her answer has always been that there is more to get done than there is capacity, which is a good problem because it means the company is doing well. AI is a capacity expansion, and specifically capacity expansion on the operations side.

Where she and the CTO were looking for real lift was customer support: the contact center, across chat, calls and email. Her caveat is that you can spend a great deal of time on this and get nothing back, so the work is picking the areas where the lift is actually available.

Twenty five integrations, none of them alike

HomeAway's strategy was acquisitive from the start. The company itself was formed by acquiring three existing vacation rental listing services, and the growth plan was to find the best listing service in each region or country and buy it. Hemmati lost count somewhere past 25.

Her role was integration, because she managed the systems.

The lesson is one anyone who has integrated an acquisition will appreciate. Every single one was different. She had been designing data models and databases for years and did not know a listing service could be modelled in so many ways. No two were the same, or even similar.

The response was structural rather than heroic. After the first few, they built a standard integration layer with templates, specified up front the elements they required, and then mapped each acquisition's data into that template. Downstream, HomeAway's corporate systems could consume the standardized output without knowing anything about the source.

Michael asks whether any integration was ever complex enough to justify abandoning the data and starting over. Her answer is that it was never an option, because these were companies already generating revenue, though they certainly talked about it. What varied was the timeline, and the variable was risk: how much could be tolerated going wrong given the size of the company being acquired, which determined how many checks and balances were needed.

What engineering gives an operator, and what it costs

Asked what her engineering roots contribute, Hemmati names detail orientation and applied logic. The if then else habit of enumerating every condition and working out how to handle each one.

Then she names the cost, which is the more interesting half. Engineers can be too committed to true or false, and she has had to learn to expand into the grays, because not everything is binary.

Her second point is about audience. Engineers want to build the next thing and move on. Operations requires thinking about the consumer of whatever you are putting in place, and whether it is genuinely valuable to them.

Nine figures in seven years

Michael notes ZenBusiness crossed the nine figure revenue mark quickly, at a company then approaching its seventh birthday, and asks what they got right.

Her answer is three things. Taking risks and testing rather than waiting for everything to be perfect. A continuous cycle of research, test, see what happens, then either build it out or move to the next idea. And staying close to customers, which they operationalized early through customer reviews and net promoter score surveys, both run through third parties.

Michael adds the external evidence: a 4.9 out of 5 rating on Trustpilot across thousands of reviews, then asks the harder question, which is how you stay close to customers when they are small businesses and there are a great many of them.

Hemmati's answer is a combination of constant customer calls, systematic feedback capture, and personal contact when someone comments on something, so they can learn what needs improving.

The example she gives is the one to remember, because it is a failure rather than a success. Over a couple of months, traffic grew sharply and customer acquisitions grew with it, and they did not have the staff to support it. The satisfaction score started falling. They knew the cause, which was response time, so the work became hiring quickly and compressing the training pipeline.

What made that possible was measuring the KPIs continuously, response time and customer satisfaction, so the drop was visible early. And, importantly, making sure everyone in the company knew how those numbers mattered, so that when something moved, the whole company reacted rather than one team.

The stack behind it: Salesforce as the CRM, a data team pulling from every system, Tableau for reporting, and analysts digging into the data to find what should be broken out next. Her summary is that data is power, and that it has been one of their largest areas of investment.

The macro signal in a formation platform

ZenBusiness sits on unusual data, because businesses both start and wind down on the platform, across a period covering a booming economy, a pandemic, a funding explosion and then a downturn.

Hemmati's read is counterintuitive and matches what she saw at HomeAway. When the economy is good, people who want to start something go do it, because the market is receptive and they can get customers quickly. When the economy is bad, people start businesses too, because they have lost a job or decided it is finally time to do the thing they always wanted to do.

Michael's summary is the right one: when the economy is up you get customers, and when it is down you get customers.

The office was the constraint

Her crazy story is COVID, and specifically what she learned from it that she did not expect.

They were about 30 people. She sent everyone home to be safe, expecting maybe two weeks. Months later, with no visible end and no certainty about the market, she decided not to keep carrying the expense, closed the office and got out of the lease.

What followed was the part she did not anticipate. Remote was working. People were happier, saving commute time and money. Some wanted to move closer to family. Some could not afford to buy a house in Austin and wanted to go somewhere they could.

Then came the realization that reframed hiring. The thing that had always constrained where they could find people was having an office and wanting people to come to it. Without that, they could hire wherever the talent was. They tried one or two more states, it worked, and they kept going. They now have employees in 30 states.

Her closing note is honest about how far outside her expectations this landed. In her whole career she never thought a company could be fully remote, not meeting in person on any regular basis, and work well.

Michael's exit line writes itself, which is that presumably they used ZenBusiness for all 30 state registrations.

The 5 things I took away from this conversation

1. A public benefit corporation is an operating commitment, not a marketing one. What makes it real is that it lands in the annual planning cycle alongside everything else. Initiatives get chosen, budget gets allocated, and there is a program to run. The grant scheme exists because the structure obligated them to have something.

2. Build the integration template after the first few, not after the twentieth. Shanaz's insight is that no two acquisitions model their data the same way, so the leverage is not in getting better at bespoke integrations. It is in defining the elements you require and making every acquisition map into them. That decision is what let HomeAway keep acquiring.

3. Chief other officer is a real job. I recognized this immediately. Things arrive with no owner, the COO takes them, and the discipline is deciding later whether to keep it, move it, or hire for it. Treating that as a temporary custody arrangement rather than a permanent expansion is what keeps the role from becoming a landfill.

4. Everyone should know the KPIs, not just the team that owns them. When their satisfaction score fell because response times slipped, the reason the company could react fast was that the whole company understood why those numbers mattered. Metrics that only the owning team watches produce slow responses.

5. The office was the hiring constraint all along. Shanaz did not set out to build a distributed company. She closed an office to save money, noticed people were happier, and then realized the office had been quietly deciding which people she was allowed to hire. Going from one city to 30 states came from removing a constraint nobody had named.

FAQ

What is a public benefit corporation? A corporate structure, most commonly established in Delaware, in which the commitment to a public benefit is written into the company's charter. Unlike a values statement, it obligates the company on an ongoing basis, which for ZenBusiness means annual planning includes initiatives specifically aligned to that commitment.

What is the difference between a public benefit corporation and a B Corp? A public benefit corporation is a legal entity type established under state law, which changes what the company is chartered to do. B Corp is a certification awarded by a third party, B Lab, based on an assessment of the company's practices. A company can be one, both, or neither.

What does a public benefit corporation actually have to do? Maintain activities and programs consistent with the stated benefit rather than declaring an intention once. In ZenBusiness's case that includes charitable giving, employee volunteering across the states where they operate, and a grant program awarding $5,000 to a handful of customers each month.

How do you integrate a large number of acquisitions? Hemmati's approach at HomeAway was to stop treating each one as bespoke. After the first few integrations they built a standard integration layer with templates, defined the required data elements up front, and mapped each acquired company's data into that structure so the corporate systems downstream could consume it without modification.

Can a COO come from an engineering background? Hemmati moved from CIO to COO at ZenBusiness and retains infrastructure engineering and data science alongside the general and administrative functions. She credits engineering for detail orientation and rigorous enumeration of conditions, and notes the adjustments required: getting comfortable with situations that are not binary, and thinking about the consumer of what you build rather than just building it.

Also mentioned

• ZenBusiness, the formation and compliance platform, and its customer grant program
• Delaware's Division of Corporations, where most domestic entities and public benefit corporations are established
• B Lab, the certifying body for B Corps, a distinct thing from the PBC entity type
• HomeAway, its roll up of vacation rental listing services, and its eventual acquisition by Expedia
• Trustpilot, where ZenBusiness holds a 4.9 rating across thousands of reviews
• Net promoter score, run through a third party from early in the company's life
• Salesforce and Tableau, the CRM and reporting layer behind the KPIs

Listen to the full episode

Shanaz Hemmati on Between Two COO's

Between Two COO's is hosted by Michael Koenig. Subscribe on Apple Podcasts, Spotify, or wherever you listen.