Sales from a buyer’s perspective

As I’ve said before, I am not a sales CEO. I’m not even much of a customer-facing CEO. But I’ve been on the other end of a lot of sales pitches, both from our competitors when I was a practicing lawyer and more generally.

Here is how I think about effective sales.

The fundamental rule of effective sales is that solving my problem is your job, not mine. You are talking to me because I have a problem to solve. I want to hear only those things that help me decide whether you are the company who will solve that problem for me.

Everything else is annoying. When there is too much “everything else,” the annoyance becomes anger. Sometimes I will buy a product even when sales doesn’t do a good job at showing me why their company will solve my problem; but in those cases, I am buying in spite of the sales effort (i.e., because I can see for myself that the product or service will solve my problem), not because of it.

So where do salespeople often go wrong?

— No deep understanding of the product. If you have me on the phone, it is infuriating to hear that you need to rope in other people (generally sales engineers / technical folks or domain experts / integration folks) to answer my questions. My first real-time interaction with you should be with someone or some group of people who can answer any question I am likely to have.

— No deep understanding of the domain. I don’t want to explain to you what I do if that is something you already should know. Nor do I want to hear you tell me about what I do when it is plain that I already know what I do. And I certainly don’t want to spend the first 15 minutes confirming with you that what you should already know I do is in fact what I do. Put another way: I don’t want you to learn about me on the call. I want to know, from what you say and what you show and how you act, that you are like me, that you understand me already. Show this, don’t tell me it’s true.

— No deep understanding of the space. I am looking at alternatives. To help me pick intelligently, you need to compare and contrast yourself to those alternatives. The alternatives are not only other products, but solutions other than products; in greenfield sales, you need to compare and contrast yourself to the status quo. And you should be giving me new information, not obvious things I’ve already thought about, when you do this.

I don’t (want to) buy on features or specs. What I am really after is not a list of features or technical specs or even a discussion of how particular features or specs will help me solve my problem. I do not want to be in the position of asking hard questions about these things (will it do X? will it do Y?) because I might miss something; you know what questions to ask much better than I do. Feature discussions and social proof (logo slides) are second bests or means to an end.

What I want is confidence that you’ve thought these things through and solved them correctly, by which I mean “in the way I would want them to be solved if I had all the information.” I want to trust that you know what you’re doing and what I’m doing and that what you’re selling will solve my problem. I then want your company to stand behind that trust and deliver something that just works.

Solving my problem is your job, not mine.

The Inventor’s Spirit

Values are why you do things (and why you don’t). Vision is the way you want the world to be. Mission is how you make the world that way.

The best values are emergent. Like a common-law rule that emerges from the decisions of judges in dozens of cases, all dealing with different facts, values are the why that unites a company’s actions in all the different situations its employees face. Why tackle this kind of problem rather than another? Why hire, retain, or fire this employee? Why arrange compensation plans this way and not another? How much should be invested in customer support and success? What should the response be to this action by a competitor? The best values are those that emerge from the actions a company takes in all these situations.

Companies can and should think about the values that their actions evince. They should identify, in writing, what they are; having done this, they can help spread those values (especially to new people during rapid growth) and can even tweak them upon reflection. This is like a judge who summarizes dozens of cases into a rule and refines the formulation of the rule in the context of the facts of the case before him, knowing that his restatement will guide future judges deciding similar cases. On one prominent model of judging, what judges do is exactly this, all the way down, in hard cases, to values that sound moral philosophic rather than legal.

Codifying, promoting, and refining emergent values is possible; legislating new ones out of whole cloth is not. You can’t command people to change the why that makes them do or not do certain things; you can only distill, emphasize, and nudge a little bit that why. This is why starting with a group of people and a way of acting that embodies the kind of values you want to end up with is so important in starting a company (or any other group of people).

Last year at Disco, Mike Wilson, who has been with us since the beginning, codified, over a series of conversations and many drafts, our values. They are why we do things: why we enter the markets we do; why we build the products we build; why your interactions with us are the way they are; why the people on the team are who they are. They are our why.

We named these values “The Inventor’s Spirit,” and you can see them listed, for example, on the team page of our website.

The Inventor’s Spirit

Observe sharply. The first thing to do is to look hard at what’s going on. Don’t accept the world the way it is; don’t assume that others have looked hard already.

Spot the issues. Identify the problems, both with the status quo and that a proposed solution is likely to create. The issues are what you are looking hard to find.

Solve problems from first principles. Don’t limit yourself to, or settle for, incremental improvements. Throw away all constraints and think of the way the world should work. Build a proposed solution from first principles, top down, from the way the world should be, instead of bottoms up by a lot of little improvements to the status quo.

Have courage to pursue new and better ways. Often this way of thinking leads you to something radically different from what came before. Don’t be scared by this; embrace it. Magical solutions often require radical change.

Important problems, impactful solutions. If you’re going to go to the trouble of solving problems this way, then you should be solving important problems. Solving important problems from first principles leads to impactful solutions, which is why this place is worth your life’s work.

Ambitious goals, humble methods. We set ambitious goals, like helping the law deliver on its promise of reasoned rule, but we pursue those goals using humble, data-driven methods. Do what is proven to work. Experiment always. Listen to what the world is telling you.

Great people + big problems = great achievements. Great things come from a simple recipe: gather the best people and aim them at the biggest problems with a license to work magic. Working at Disco isn’t easy; we do it because we think the outcomes are worth everything.


Occasionally you read or hear an idea that sticks with you, that makes a big difference in what you think and how you act. Often the idea is obvious in hindsight. Some examples for me are the ideas of the law lord as a model of the law, the Necker cube as a model for storytelling from evidence and what lawyers do, and the following idea about persuasion.

The easiest way to persuade someone to do something is to show that it will get them something they already want. For example, if you’re trying to convince a developer to develop in a way that is minimally harmful to the environment, and you know the developer cares about money and not about the environment, you should show the developer how doing the environmentally friendly thing will make him more money. This might be because not doing the environmentally friendly thing will lead to expensive litigation down the road or because doing the environmentally friendly thing will win him support in the permitting process or for any number of other reasons. It is much less effective to try to show the developer that he should care about the environment.

This is true regardless of what you think about the developer’s valuing money over the environment and it is also true regardless of what you think about the arguments why doing the environmentally friendly thing will make the developer more money. What matters is not what you think of these arguments, but what the developer does. And this is true even if the developer knows that you don’t care about his making more money; the developer does care, and that is enough for him to assess your arguments on the merits and be persuaded by them if they are otherwise persuasive. If your goal is to convince the developer to do the environmentally friendly thing, then you need to link doing that thing to what the developer already cares about.

Another way of putting this is that it’s generally easier to convince people about means than it is about ends; it’s easier to convince them that they’ve misunderstood the best way to get what they want than to convince them to want something else entirely. It is not intuitive (at least it wasn’t for me before hitting the idea) that this is the way to go, especially when you yourself care deeply about the ends in question and think someone who doesn’t share those ends is deeply misguided or evil. But what you care about doesn’t matter; what they care about does.

I remember hitting this idea for the first time in some book in college; I don’t remember what book it was, but both the idea and the specific example are taken right from it. If I can teach you only one thing about persuasion — in front of juries, in sales, in recruiting, in finding witnesses, in fundraising, in any area where you want to convince someone to do something — this idea is it.

$10M in funding for Disco

We are delighted to announce that we have closed a $10 million Series B investment led by Bessemer Venture Partners with participation from our existing investors LiveOak Venture Partners. Bob Goodman, founding partner of BVP’s New York office, has joined our board. You can read the official press release here.

In 2013, we launched the first version of our ediscovery product, won our first customers, and closed our first round of venture capital. We were truly a bootstrapped team: two engineers (Gabe Krambs and Barry Hagan); one salesman (Alex Torry); two lawyers doing everything from customer service to sales meetings to giving CLEs (Kent Radford and Mike Wilson); my longtime assistant (CeCe Cohen); and me. In retrospect, it’s amazing how much this small team was able to achieve. We would not be here without their hard work, and the sacrifice of the family members who let them give their all to Disco.

2014 was a year of experimentation and refinement: major improvements to our product infrastructure; much-demanded features to improve the user experience for lawyers; deeply understanding our market, customer segmentation, pricing, and sales strategies; putting a basic channel-partner program and channel support in place; and developing our first real marketing collateral.

It was also a year of growth.

In engineering, we added Jason Vos, Daniel Posada, and Levi Bucao to the infrastructure team led by our CTO Barry Hagan and created a UI team with Kalley Powell and Brian Garcia led by Lead Engineer, UI Brian Ng. In sales, we welcomed long-time Compaq and HP executive Tom Benwell as our VP Sales and created a system of inside sales associates in Houston (Daniel Donnelly, Wheatley Gates, and David McMillian) and regional sales executives in territories around the United States (Alex Torry, Mike Wilson, Michael Bender, Dustin Whipple, Keith Mobley, Robert Schaffner, Bob Fogarty, and Stephen Mace) held together by sales coordinator Kelly Neuner. We formed an operations department led by Lead Engineer, Operations Gabe Krambs with Julia Englander, Ross McDonald, and more soon to follow. And we brought on Steve Kahan, Houston’s winningest startup VP Marketing (six exits, including Postini to Google), to build a real marketing group at Disco. Our team was 7 at the end of 2013; we are 27 today.

Customers received Disco warmly in 2014, with our users now including lawyers at more than 300 law firms, enterprises, and government units, including 40 of the AmLaw 200. With more than 20% month-over-month recurring revenue growth, Disco is now the fastest-growing technology in ediscovery. Two of our customers loved Disco so much they invested in our Series B, one a senior lawyer from an AmLaw 200 firm and the other the former general counsel of one of our first corporate clients. And we have been delighted by the number of end users and channel partners who volunteered to film video testimonials with us, testimonials we will be sharing over the next several months. To the Disco team that made this growth possible: thank you.

2015 will be a year of scaling. With sales and marketing leadership in place and a real sales team ready to go, we will replicate the success we’ve had in Texas and other initial markets across the United States. If we do our jobs correctly, every litigator in the United States will have heard about Disco and our automagical, do-it-yourself ediscovery model by the end of 2015. Our engineering and operations teams will also grow rapidly to maintain our technology lead, especially on speed, and to ensure our customers continue to receive first-class service directly from the people who built the product.

As a company, we will also begin to publicly articulate the vision that brought us all together, our vision of a world of law done right.

We’ll have more to say about this, both on this blog and in person around the country, in the coming months. But our mission as a company is not simply to build a great ediscovery product; it is to systematically automate all the parts of the practice of law that don’t require independent legal judgment so that great lawyers can focus on doing only the things that only they can do. The time has come for legal technology that empowers great lawyers and great law firms to deliver on the promise of the law to clients and society.

Disco is the company at the intersection of engineering and law. We unite world-class engineers with a deep love and respect for the law. Lawyers, including former partners and associates at some of the most respected law firms in the world, are present in all departments at Disco, including engineering, operations, and sales. With this team, we are confident we have reinvented ediscovery; over the next decade, we will reinvent law.

I am, every day, humbled and honored to serve as Disco’s leader as we set out to change the world for the better.

Tips for raising a Series B

Last year, when we raised our Series A, I wrote Reflections on Raising Venture Capital to recap our experience for the benefit of future startups. I benefited immensely from the material I read before raising and wanted to contribute back what we had learned.

This post, in the same vein, highlights some things we learned in the course of raising our Series B. Unlike the Series A post, I’m not going to include a full narrative of our raise, but only some of the highlights and the most important takeaways.

We wound up raising just over $10 million in a round led by Bob Goodman at Bessemer Venture Partners. BVP is one of the top-tier VC funds, with 111 IPOs, including 11 in the past four years. Its successes include Cornerstone, Broadsoft, LinkedIn, Yelp, Millenial Media, Eloqua, LifeLock, Bladelogic, Gartner, Vertica, and Endeca and its portfolio includes Pinterest, Box, Twilio, and lots more.

(BVP is also famous for its public “anti-portfolio,” companies it could have invested in but passed on, which includes Apple at a $60M valuation, eBay, FedEx, Google while it was in a garage, Intel, Intuit, Lotus, Compaq, and PayPal. Ouch!)

We had our first call with BVP on October 3, and the round was closed and funded on November 17, 46 days later. This compares with 48 days from first meeting to closed and funded for our Series A.

Here are the tips:

Warm intros were not more effective than cold outreach or VC inbound. The common wisdom is that the way to raise money is to secure warm introductions to the appropriate partners at venture-capital firms. Many prominent VCs have publicly stated that they do not consider submissions “over the transom,” i.e., without one of these intros. This is a nightmarish requirement for those of us who are not already well connected and who hate schmoozing. In my Series A post, I observed that contrary to the common wisdom, all our term sheets came as a result of completely cold email outreach. Likewise, the two term sheets we seriously considered for our Series B came one from cold outreach and the other (BVP) from inbound from the VC. While warm intros from our existing investors this time around certainly converted to meetings at a higher rate, we did not find warm intros to be either a requirement to doing a successful round or even more likely to convert into term sheets.

If you commit to a raise, it will monopolize all your time as CEO. In my Series A post, I explained how we were ambivalent about raising money and turned profitable during the raising process. This meant that raising money was actually not much of a distraction and really more of a background process for me. This time, we committed up front to doing a Series B to fuel rapid growth (although we did arrange a backup bridge round with our existing investors), and, as is the common wisdom, it took 100% of my time during the raising process. For roughly two months, I was able to participate only minimally in engineering, operations, sales, finance, and other essential company functions. If you are a very hands on CEO like me, this is not fun.

Lawyers affect the experience, but not the outcome; push, push, push on the other side’s lawyers; do not assume they are working if you aren’t pushing. The style (not necessarily the quality, as all the lawyers you will deal with for major funds are baseline good) of the lawyers greatly affects how pleasant the document-drafting process is, but it does not affect the bottom line outcome in terms of financials or control. One practical tip here is not to assume that the lawyers are quietly working on documents; instead, you should continuously push to ensure that things are moving quickly to a conclusion. Lawyers who are working hard on your deal will turn documents daily. Delay beyond one day between turns of drafts between your lawyers and theirs is caused by either the deal not being a high priority or unavailability of someone on the other side (read: usually, the same thing). If lawyers are instructed by their clients to go quickly, this speed is definitely achievable. The trick is to do this while still appearing pleasant to deal with (I probably erred on the side of being too pushy, but I recommend doing the same).

Enlist a great UI guy to work on the pitch deck. VC run-throughs are invaluable. If you are a SaaS company like us or otherwise have beautiful end-user-facing software, then you have brilliant designers on staff working on your UI. Use them to help you with your pitch deck! I am shocked in retrospect at how ugly our Series A deck was and how beautiful our Series B deck was once our head of UI agreed to help me on its design. As with all design, this is not merely about polish or surface appearance; great design makes the ideas flow better and makes the story more persuasive. Don’t neglect this. Similarly, in designing the content of the deck, run-throughs with your Series A investors are invaluable; we made many major changes (including adding something called an “innovation and incumbency chart,” which is apparently a McKinsey thing that I found at first blush absurd but that resonates with MBA types) after a first run-through with our investors that we would not have come up with had we not done the run-through.

If you are a VC, the personal touch matters deeply. Relatedly, once you decide you want a deal, don’t negotiate unless it really matters. Lots of personal contact between the partner leading the deal and the founders is the best way to make your firm attractive and to make the founders comfortable with you. You can jump way up the list (over “more prestigious” firms and even over firms making better financial offers) by doing this. A corollary is, once you decide to do the deal, stop negotiating over stuff you don’t deeply care about. It is better to get the deal at slightly worse terms than to snatch defeat from the jaws of victory by trying to negotiate in the endgame. If a founder tells you what will get the deal done and you want to get the deal done, the safest answer is, “Yes, we accept. Here are the definitives.” If you want to tinker with things later in a way that you and the founders think really is win-win, you can do that later; but first, lock in the deal.

Round sizing is hard, especially outside the Valley. A small round by Valley standards is a big round by Texas standards. So it is very hard to know what to ask for. I don’t really have a good solution here, other than to flag the issue. If your baseline sense of what is normal is set outside the Valley, you should probably raise much more (2x or more) than you think you need or than you think is appropriate, not so much out of conservatism, but because, even knowing this and trying mentally to take it into account, you will likely be wrong. If you are a first-time founder like me, and you are not steeped in Valley valuations, just remember to adjust sharply upward. It is easier to move the ask down than to move the ask up.

Negotiate control and financial terms in that order; everything else is much less important. This is what I said in context of the Series A. It is also the conventional wisdom. It is correct. Retaining control is most important. Financial terms are next most important. Everything else is stuff for the lawyers (and you should try to keep your lawyers from slowing things down by fighting about stuff in this “everything else” category).

If some customers are great for you and other customers are bad for you and it is hard for VCs (or perhaps even you) to tell which is which, then luck of the draw during VC diligence will strongly affect your outcome with particular funds. We sell ediscovery software that lawyers use themselves. This contrasts with all other players in our space, who sell ediscovery software aimed at litigation-support or IT staff who work for lawyers. In sales, this means we have a magical close rate when showing a demo to lawyers, and a horrendous close rate when showing a demo to litigation-support or IT staff (or to lawyers who are just there because a VC asked them to go to the meeting, but really they defer to their staff). Because of this, some VCs had awesome experiences in diligence (everyone they showed the product to loved it) and some VCs had awful experiences in diligence (they could not find a single person who would buy it). Obviously the first category moved forward and the second did not. If your market is similarly bifurcated, you will experience this frustrating “luck of the draw” effect while raising. You can explain what’s going on to the VCs, but they will naturally trust the market feedback more than anything you say.

VCs who have had bad experiences in the space are going to be out (and this is the only thing we encountered in either raise that trumps growth numbers). Two funds that were initially enthusiastic backed off because of a bad prior experience in our space. One of them had a partner who was the CEO of a startup in our space; the other had a partner who was the CEO of the company that acquired that startup. While the startup had a good outcome, it was a slog for them. This experience caused both former CEOs to be strongly negative and caused both funds to pass on the deal. This competitor did more harm to us this way than it ever has in actual competition for customers!

Collect all diligence materials up front and have an online document room to make sharing them easy. We did the former, but not the latter. We might have been able to shave a week or so off the schedule by doing the latter. Not a big deal, but super easy to do and something I would definitely do next time around.

* * *

These are the things I wish I had known before raising, so I hope they are helpful to those who come after. Houston-based startups, feel free to reach out to me if you want to discuss any of this in more detail.

Defensibility in Disco

Disco provides end-to-end defensibility that lets you audit the path of every document from when it arrives in Disco through production and eventual matter deletion.

  1. New data is tracked in FTP logs for data transmitted by FTP and using chain-of-custody forms for data on physical drives delivered to the Disco office.
  2. Each ingest session generates a complete ingest report that indicates the disposition of every file from that ingest session. The report says what was done with the file (ingested, deduplicated, deNISTed, etc.) and reports any ingestion errors (password protected with no password supplied, etc.).
  3. Audit logs contain every action taken in Disco during review, who took the action, and when it happened, all with full granularity. This includes, for example, every search that was run and who ran it, every document that was viewed and who viewed it, and every tag that was applied or removed, when it was applied or removed, who applied or removed it, and whether it was done individually or as part of a mass tag.
  4. Production indexes and SHA1 and MD5 hashes of production zips ensure integrity of productions.

With FTP logs and chain-of-custody forms, complete ingest reports, audit logs, and production indexes and hashes, you can audit and defend each step of the ediscovery process.

LFD 10 – Sensible defaults

(This post is part of the Lawyer-Focused Design series, which explores 10 ways in which Disco is designed for lawyers.)

Sensible means that you can almost always use the default option in Disco and Disco will do the right thing; defaults means that when you do need to do something else, you can.

For example:

  1. When you tag an email, the tags flow down to the email’s attachment. This is sensible because when you produce an email, you generally mean to produce the whole thing. But, when you need to, you can remove the tags from some or all of the attachments or apply different tags to them altogether.
  2. When you produce documents, Disco produces one PDF per document with an industry-standard EDRM format load file and one copy of each document per custodian and per family. This produces documents by custodian (that is, in the manner in which they were kept) and produces emails complete with any attachments tagged for production (just as you would get if you printed the emails out). This is normally 8what you want. But when you need something else — a production that complies with the SEC’s archaic production rules, or a production with a custom sort order to comply with a 26(f) agreement — Disco can do that too.
  3. When you produce an Excel file that doesn’t image well, Disco produces the native — unless you’ve redacted the Excel file, in which case Disco is smart enough to produce the redacted images. And whenever you apply redactions, Disco knows to replace any OCR or extracted text with new OCR run on the images so that no information leaks to the other side. When you need control over exactly what gets produced as native, or with a native, you can drill down and make those decisions; but 95% of the time, you can trust that Disco makes them for you correctly.
  4. When you’re reviewing documents, you see deduplicated documents so that you don’t review the same document more than once. But, when it comes time to produce, you can select the level of deduplication, or produce only some instances of deduplicated documents (for example, the ones that belong to a particular custodian or that were found on a certain computer).

Disco takes care of the ediscovery details so that you can focus on winning cases, not tinkering with file formats, deduplication levels, or load files. But when you do need to worry about the details, all the power you need is one click away.