Own More of Less

The Core Insight

In January 2011 the chief executive of HubSpot offered Rand Fishkin 20,000,000 to 30,000,000 dollars for Moz. Fishkin asked for 40,000,000. The reply the next morning was 25,000,000 dollars, and he turned it down.

Moz was on track for about 10,000,000 dollars of revenue that year. Fishkin held 32.5 percent, so the offer was worth 8,125,000 dollars to him. He has re-read that arithmetic fifty times in the six years since.

Lost and Founder is the ledger of what happened instead. Moz reached 42,000,000 dollars of revenue in 2016, 7.5 times its 2010 number. He and Geraldine hold about 23 percent of it now, behind a liquidation preference of one times the 29,100,000 dollars raised. The refused offer was cash and stock, and that stock rose sharply after the 2014 offering. A 250,000,000 dollar exit does not clearly beat it.

Most founders treat a venture round as the first real milestone, and read the stories of successful companies as instructions. Fishkin reads those stories as the reporting of survivors. The financing model behind them is built for outliers, and its terms only fit a company that can be one.

The base rates sit in his introduction. Harvard Business School cohort work puts the share of early-stage technology companies that fail to return investor capital above 75 percent. Restricted to venture-backed technology startups, that figure passes 90 percent. Half the companies still alive in year four went under.

The Framework

Seventeen chapters each open with a quote from a startup luminary and then take it apart. Five claims carry the argument.

• The advice founders inherit describes the exception, because the companies that failed never wrote it down.
• Venture capital needs outliers, so its terms punish the company that turns out merely good.
• Founder ownership decays while revenue grows, so a bigger company later pays the founder less than a smaller one now.
• Focus is the growth variable, because every product added subtracts from the growth of the others.
• Founder weaknesses become permanent company debt, and hired-in strengths leave at the two-year mark.

Fishkin can argue from a ledger because Moz published one. The company put its financials, its failed fundraises and its product failures in public, and was called foolish for oversharing. Transparency is saying much, and honesty is saying only what is true.

Revenue in the printed table runs from 800,000 dollars in 2007 to 47,400,000 in 2017. From 2012 to 2016 Moz consumed more than 35,000,000 dollars of venture money and debt while growth fell to about 10 percent. The venture world treats 30 percent as the floor for an interesting business.

The audience that transparency built shows up in the funnel. A visitor who arrives from Google search and signs up at once stays less than four months, against a global average near nine. A visitor who came twelve or more times in three months first stays past fourteen.

Key Ideas

The Fund Math Fits Almost No Company

Ignition Partners raised 300,000,000 dollars in 2004, owing its limited partners 900,000,000. In November 2007 it put 1,000,000 dollars into Moz at a valuation of 7,100,000, for about 14 percent.

Run the exit. At a 40,000,000 dollar acquisition, Ignition takes 5,600,000 dollars, a 5.6 times return delivered early. That outcome is useless to the fund. Returning the fund on those terms takes three hundred companies performing exactly that well.

A 400,000,000 dollar fund owes 1,200,000,000, so a sale returning 112,000,000 on 15,000,000 invested still fails it. The investor is paid to block the exit that makes the founders rich.

Of ten investments, five fail, three return small amounts, and two produce most of the gains. About 5 percent of venture firms beat the market, and the bottom half return less than the fund they raised.

Ownership Decays While Revenue Grows

Moz walked into the HubSpot meeting with 5,700,000 dollars of 2010 revenue, net margin near 80 percent, and growth around 100 percent. The offer came to 4 or 5 times revenue, where SaaS deals ran 4 to 10 times. Investors made back three to four times their 1,100,000 dollars, so the refusal was reasoned.

The refusal aged badly for one reason. Revenue multiplied 7.5 times while the founding share fell from 32.5 percent to about 23. The average successful startup raises about 42,000,000 dollars, against 29,100,000 at Moz, so the dilution was not finished.

In 2011 twelve early employees held 0.5 to 3 percent each. At the 25,000,000 dollar price that is 125,000 to 750,000 dollars apiece, at a near-zero strike. A comparable employee in 2020 holds 0.05 to 0.1 percent against a common stock valuation of about 65,000,000 dollars.

A Minimum Product Costs More Than It Saves

Moz Analytics came from a theory Fishkin held in 2011 and never tested outside his own head. It shipped in November 2013, after five slipped dates and a replaced engineering lead. Of the 90,545 people who signed up for launch notification, 2.3 percent paid for even one month. Matching the old product took almost another year, and the damage ran three growth-stunted years.

A week-long job swap with Wil Reynolds of SEER Interactive killed the all-in-one thesis. Consultants verify every data point by hand before trusting a tool, and they switch tools with no loyalty and no switching cost.

Spam Score is the cleanest failure. About a hundred candidate factors were cut to seventeen flags, built by five people over three months spread across a year. Six months after launch, about 5 percent of Open Site Explorer users had touched it, with no observable effect on trials, retention or growth. The research and engineering behind it cost at least 500,000 dollars.

His replacement rule fits a line. Build the minimum product, launch it privately, and release only the exceptional one in public. The damage he prices is a brand memory that outlives the fix.

Keyword Explorer is the proof. The launch slipped from year-end to January, and one critical demo bought four more months of work. It landed nearly six months late and drew more than seventy thousand visits in two days. At 149 dollars a month inside the subscription the feature sold. Sold alone at 600 dollars up front, it found almost no buyers.

Focus Is the Growth Variable

On August 17, 2016 Moz laid off 59 of its 210 full-time employees and shut two products. Headcount had gone from about 125 at the start of 2014 to more than 220 two years later. By mid-2016 the company sold eight separate things.

The board meeting booked for three hours ran seven, with twelve months of operating cash in the projections. The cut plan targeted 12,000,000 dollars and found 12,810,000, of which 8,800,000 came out of people.

He prints five causes, and the market is not one of them.

• Retention lost to acquisition: average subscriber tenure was eleven months in 2016, against 120 months at Salesforce.
• Extra products diluted the brand: fewer than 5 percent of more than twenty-five thousand Moz Pro subscribers used Followerwonk monthly.
• Priorities scaled faster than headcount, so the third and fourth products cost far more than the second.
• Competitors passed Moz on four of six core SEO functions within four years, and one ranking position justifies switching tools.
• The law of large numbers turned 4,100,000 dollars of 2016 growth into an 11 percent disappointment.

The fix for the first cause arrived years late. A customer success call of thirty to forty-five minutes made subscribers stay 30 percent longer.

After the cuts, marketing went from twenty-five people to fourteen on a budget cut by two thirds, and traffic and free trials still improved. November 2016 was the first cash-flow-positive month in four years.

Founder Traits Become Permanent Company Debt

Moz was founded by marketers, so engineering was its permanent weakness. The web index the whole link product depended on launched in October 2008. Over five years more than a dozen hires worked that one project, and Moz still lost link-data leadership to Ahrefs and Majestic.

The mechanism is tenure. Average startup employee tenure runs about two years, so a strength you hire is rented. A founder trait stays for the life of the company. The repair is a founder who learns the weak function instead of being shielded from it, and Sarah Bird did it in eighteen months.

The hiring mistakes have the same shape. After the 18,000,000 dollar Series B in 2012, twenty-five of about forty non-management employees asked to become managers. Google's own list of eight manager behaviours puts technical skill last. Moz built a real individual contributor ladder.

The costliest hire was a senior one who clashed with the team, was promoted anyway, and made sexist remarks nobody reported. The head of HR had heard it from no one else, and the people who saw it did not expect reporting to change anything. The fix was a hiring screen that gives an unrelated team a veto.

Severance in the layoff ran six weeks minimum regardless of tenure, plus a week for every year of service with no cap. It was paid as a lump sum so people also claimed unemployment. That cost 1,400,000 dollars, about 20 percent of the cash left, and Fishkin threatened to quit twice to keep it. The board approved it anyway.

Founder Depression Is the Median Case

In September 2014 Brad Feld put one question to a room of portfolio chief executives. He asked who had experienced severe anxiety, depression or another mental disorder while running their company. Every hand in the room went up but two.

Fishkin dates his own symptoms to 2013 and the name for them to the summer of 2014. He berated executives and investors, and talked his own subscribers into switching to competitors from a conference stage. Sleep went first, and fewer than 1 percent of humans function on under six hours a night.

A call with Brad Feld in November 2013 led to Sarah Bird taking over as chief executive in February 2014. Fishkin became an individual contributor.

The transferable part is a rule for getting out. Invest in behaviours, not outcomes, because judging a behaviour by its outcome makes you drop everything that does not pay at once. He kept the anti-work night, physical therapy and exercise, and dropped meditation, acupuncture and the apps.

Practical Applications

Run the ownership arithmetic before the round rather than after an offer arrives. Median founder ownership at exit is about 11 percent.

Price the services path against it. Of two 10,000,000 dollar companies, the one that raised nothing sells for 15,000,000 dollars and its founder keeps all of it. The one that raised 8,500,000 dollars sells for 40,000,000 and its founder keeps 6,000,000. Services firms pass year five at 47.6 percent, against under 25 percent for tech startups.

Validate before you build. List roughly a hundred likely buyers, interview each on how they solve the problem today, and count the signups on a teaser page.

Count what you sell. The second product was cheap to add, and the third and fourth were not. Measure retention on the first before you add anything.

Who This Is For

The founder who gets the most from this book has not raised yet, or has raised once and is being told to raise again. It is a business software book, and the company behind it ran gross margins above 75 percent.

Skip it if your business only works at outlier scale. The venture instrument exists for those companies, and the arithmetic here is aimed at everyone else.

Read it as one founder's account of one company that did not reach the outcome he wanted. The figures are Moz's own reporting, unaudited, and a few of them do not reconcile across chapters. His conclusions about venture capital are shaped by his own experience of raising 29,100,000 dollars and being unable to sell. The counterexamples on this shelf are real, and the same instrument that trapped Moz built companies with no other route to existing.

The Decision

Run the test on your own funding assumption this week. Write down the price you privately expect to sell for, and the share you expect to hold when it happens.

Multiply the two, then subtract the preference stack sitting ahead of you. Compare what is left against the best offer you can get in the next twelve months. Then compare it against twelve years of the salary you gave up.

If the venture path wins that comparison by a wide margin, raise. If it wins by a little, you are paying a decade of control for a rounding error.

Fishkin did that arithmetic six years after the offer. It takes an afternoon to do it first.