Jan 07

getting to plan BI just finished the book Getting to Plan B. This is hands-down the best book for a startup I’ve read in the past year. It’s got just the right balance of theory anchored by real-world examples to make the lessons stick. It’s like basic anatomy for business and it shows you how to think about the all the organ systems involved, how they function together and how tweaking one ripples effects to the others. I’ll a do a brief review/summary here for anyone who is thinking of reading it, and if you’re in a startup this should be the next thing you read.

Synopsis

The core tenet of the book is that business planning is a misnomer, it’s more accurately business guessing. And companies almost never pick their destiny right on the first guess. The art of building a startup is the iterative process of zeroing in on the right formula over time. It’s all about making and testing hypotheses with empirical data rather than drafting a massive plan up front and clinging to that. If you’re familiar with software development methodologies this is the age-old “waterfall vs. agile” distinction. And this book is essentially the Agile Manifesto of business formulation that provides a creed as well as a framework for how to think about your business and conduct this iterative development.

The book is divided into two sections: the verbs and the nouns. The verbs are the processes you’ll use to hone your plan with each iteration. The nouns are the five models or “organ systems” to consider as you look at your business. First let’s look at the verbs:

The Verbs: Processes for refining the business

The ultimate goal is to zero in on the best formula for doing things that lead to the best outcome as quickly as possible. I’ll use Netflix (before it existed) as an example to demonstrate each concept.

  • Find Analogues: These are the parallels you can look to in order to answer questions without doing anything other than research. The idea here is to not reinvent something that’s already been proven. Find relevant design patterns from existing businesses either in your space or better yet, in a completely different industry (because competitors likely will have already considered your adjacent players). Glean what you can from what these guys have already proven and treat it as free brickwork in building your foundation. In the Netflix scenario this might have been looking to businesses like Columbia House or BMG for proving that people were willing to purchase media mail order via subscription service.
  • Find Antilogues: These are the anti-patterns to study- either the failures that have come before you or the businesses that are out there succeeding marginally but doing something wrong. Study their shortcomings and resolve to consciously do things differently. Again, this is freebie insight you get from analyzing the efforts of others – these bricks cost you nothing and allow you to work from the base of experience others have proven out rather than starting from dirt. In the Netflix scenario Blockbuster could be an analogue for its physical stores, per rental fee & late fees aspects. Another one might be Divx for how it attempted to rent DVD’s and then expire them using DRM technology. Basically anything that’s been tried and either failed or shows status quo thinking and can help you decide what not to do.
  • Identify your Leaps of faith: These are the important questions left over once you’ve applied the relevant analogues and antilogues- basically the stuff that keeps you up at night wondering if it will work. Presumably you’re going to be breaking new ground- these are the do-or-die questions that will determine the viability of your proposed business. The success of your plan hinges on proving or refuting these questions as soon as you can. In the Netflix scenario some key leaps of faith: A) would people ever embrace the mailed DVD approach or would they need the spontaneity of in-store rentals? B) Would the turnaround times be sufficiently quick to provide enough liquidity and selection in the catalogue? C) Would loss or damage from postal mail as the transfer medium add costs making the service unfeasible under rates the customer would accept?
  • Test them via Dashboards: These are the collections of metrics you monitor to prove or disprove your leaps of faith. For each leap of faith you determine the specific metrics necessary to test the validity of your hypothesis. For the Netflix LOF’s above some dashboards would have been: A) adoption as measured by signup rate, random surveys to determine % of movie watchers who used Netflix vs. traditional rentals over time B) avg time users held a movie, satisfaction ratings on selection quality, avg time movie was in transit C) user complaints related to loss or damage, inspection reports upon receiving returned movies.

The Nouns: The five models to consider

If you distill it all down, there’s five aspects of your business you need to refine. They’re all interrelated and tweaks to one will impact how the others function. None of this is rocket surgery when you examine them in isolation but it gets interesting when you see how changing one can allow you to play with the others for strategic advantage.

  • Revenue Model: Who/what/when/where/why/how will people buy your stuff? How often? At what price? On what terms? Can you think of more meaningful ways to sell the same products under different assemblies that will increase the perceived value and allow you to charge more?
  • Gross Margin Model: All about cost of goods vs. revenue – how low can you drive your COGS and how high can you kite your prices? What are the knobs you can twist on how you spend on packaging, manufacturing, delivery and sales to help your gross margin approach 100%? How can you hedge in scenarios of uncertainty using multiple product lines with mixes of different gross margin models?
  • Operating Model: How can you slim down your overhead? Is it possible to transition fixed operating costs to variable in the short term so you have the benefit of that cash early and optimize for profit later? What assumptions can you challenge and what can you cut out of your offering that is non-essential and can make you a lean athlete and give you a competitive advantage in your operating model?
  • Working Capital Model: Available cash as determined by Assets minus Liabilities. We all try to pay our bills at the last possible minute and collect our receivables as soon as possible but what are the real implications? How can being ruthless in optimizing this equation allow you to minimize investment required or make life difficult for a competitor? Can thin margins allow you to pass savings onto customers, create volume and allow you to negotiate better payment terms with suppliers giving you more cash on hand?
  • Investment Model: the cash and other resources needed to get things started, get to break even and to then grow. How can you stage rounds of investment to increasingly remove uncertainty and therefore raise the capital you need on better terms? What can be bartered, eliminated, deferred or substituted to reduce the needed cash investment up front? What culture can you instill early on by taking a spartan approach? How much of the pie can you keep in the early stages so you can splurge on giving your employees more options and retain more control in later stages?

Takeaways

Isolate the leaps of faith and develop the metrics that test each independently: We’ve been intuitively testing hypotheses and adapting our business at JumpBox based on feedback since day one. What Plan B helped me fully absorb is the benefit of unraveling knotted hypotheses that involve multiple leaps of faith and thinking about each individually. Think of some unknown in your proposed business plan with a seemingly-straightforward question like “will people buy durian-flavored lemonade?” and I bet you can decompose it to constituent questions of “will people buy lemonade that’s a) purple b) tastes like durian c) branded with a durian fruit image on the cup? This is a simplistic/silly example but the point is the more you can tease apart the variables the better you uncover the true drivers.

Think about how you can unlock more cash: The gross margin, revenue and operating models all affect your working capital model which then dictates what you need to do investment-wise. The more ways you can find to free up cash (whether by improving your gross margin, reducing operating costs, getting your customers to pay you in advance, extending terms on accounts payable, etc) the less money you have to raise and the less equity you have to give up. This of course should be common sense but the examples in the book of Costco, Go and Skype hammered these lessons home and showed how you can not only grow your own business but actually suck the oxygen out of the room for others and create impossible living conditions for your competitors.

The importance of the dashboard: We already monitor key metrics in our company but Plan B hounded importance of aggregating these figures in one place and snapshotting them over time so you can see unequivocal evidence (and substantiate it to others if necessary). Traditional business intelligence systems are too heavy for most early-stage startups. But free online tools like yahoo pipes & dapper feeding a google spreadsheet can give you much of what you need. Of course if you’re slightly more technical there’s ETL and data presentation JumpBoxes that can give you even more control over your dashboard ;-)

Sources for inspiration of analogues and antilogues I read magazines like Wired, Fast Company and Inc and I find the stories of the companies interesting but Plan B gave me a new way to think about the companies covered in these articles. It’s impossible not to start envisioning what their underlying working capital model must look like and start thinking of them as an analogue or antilogue. I love books that peel back the translucent film on life and allow you to look at something in a completely new light and see it more clearly – this book absolutely does that. It makes reading these magazines take on a new “treasure hunt” aspect and gives leisure reading a very real prospect of producing insights that can be put to work. Awesome, just awesome.

Critique

The only major deficiency I see with this book is that it hovers at a strategic level and never dives under the water to give truly tactical advice on how to do the dashboards. The dashboard is arguably the pivotal piece in all this because it’s how you determine if you’re right. The book has a section towards the end that helps you with “what do I do next?” but it never presents example dashboards to demonstrate how they work in practice. I would LOVE to see a paperback workbook complement to this novel. Even better, I’d like to see Komisar & Mullins team up with someone like MindTouch and include a CD complete with a functional piece of example software that shows a real dashboard referencing live data in spreadsheets, a site, a database, a CRM system, etc.

On a tangent, this is precisely the type of thing we want to enable by allowing an author to bundle working, data-filled JumpBoxes on CD with a workbook. I can easily envision including JumpBoxes for MindTouch Core, Snaplogic, SugarCRM and MySQL and then having them pre-filled with actual data and referencing Google Spreadsheets, a live web site and Excel files to show how an actual dashboard works in practice. Rather than talking abstractly about measuring nebulous things like support efficiency and sales conversion, it would be great to be able to see exactly how this works. Those physical examples would bridge the last mile here – the piece that’s missing which allows you to close the book and open up your laptop and apply dashboarding to your own situation. If the authors happen to read this, let’s talk- I’ll help you write this workbook and give you full working examples that anyone can use in minutes to play with a live dashboard in action.

Conclusion

A good book gives you theory backed by concrete, practical examples that demonstrate the concepts and emblazon them in the memory for later recall. A great book gives you this material but in a way where you can’t help but look around and start seeing the world differently. I found myself constantly thinking about other businesses and our own through the lens of these nouns and verbs. The biggest benefits to me have been to help disentangle my thinking where there are multiple leaps of faith wrapped upon one another. It’s also helped me clarify how the interplay of the various models and how deeply attaining the holy grail scenarios of negative working capital and 100% gross margin can liberate and propel a business.

Of all the books I’ve read on business and entrepreneurship, if I had to recommend just one it would be a tossup between this one and Innovator’s Solution. My head is spinning with thoughts and I feel like I’ve only retained maybe 30% of the material so I’m inclined to turn to page one and read it again. I’d say towards Eric Ries’s challenge of developing a working theory of entrepreneurship, Plan B comes about as close to a bible as any I’ve found so far.

Dec 31

psychicHere’s ten things I predict we’ll see in the IT/computing industry in 2010 (and yes, I’m biased about some given the world we live in at JumpBox):

  1. Self-healing applications become commonplace: We’ll see the rise of preventative and predictive technologies that fix problems in applications before they become fatal. Monitoring systems can already intelligently scale computing resources allocated to an application by detecting when it’s hitting a resource wall. But beyond this capability we’ll see a new set of tools arise that automatically intercedes and conducts repairs on the fly by reverting to a snapshot of the app and re-injecting data. This won’t be for financial applications and mission critical apps but it will happen for apps that need high availability with data that’s “good enough.” The net effect will be that the apps are perceived as being more stable when in reality the real hero is this adaptive repair technology behind the scenes.
  2. “Brick laying” in IT gets commoditized and the IT admin’s focus returns to architecture: By “brick laying” I mean the tedious, manual processes of maintaining and provisioning applications on the network. Virtual appliances deployed on private clouds will free admins from the menial chores of wedging the next PHP app onto an existing server and enable them to focus on proactive rather than reactive pursuits. Some admins will fear obsolescence and seek job security by keeping practices esoteric and arcane but the smart ones will realize their craft is merely shifting to the more interesting duty of architect with a focus on how to leverage things like virtualization and cloud computing to keep users happy.
  3. Balkanization of non-critical IT systems in the enterprise: We’ll see the proliferation of small, rogue collaborative applications in the enterprise. This will stem mainly from the frustration of being shackled by the company’s monolithic enterprise collaboration system. As self-serve deployment of collaborative apps becomes more feasible for non-technical folks the do-it-yourselfers will circumvent IT altogether and implement the apps that make their jobs easier. These transient, project-specific apps will blossom, serve their short-lived purpose and then vanish without ever involving IT. The more territorial admins will see this as chaos and try to retain control while the enlightened ones will realize that non-critical app governance is merely being pushed out to the edges where it belongs.
  4. Someone successfully addresses data interoperability amongst SaaS and local apps: As these silo’d supporting applications sprout up both inside and outside the firewall, it becomes important to have a way to share and manipulate data amongst them. Technologies for deploying the apps will have made them trivial to deploy but the connective tissue like REST and SOAP APIs will still be way too technical for the layperson to use. ETL (data Extraction, Transformation, Loading) products like Jitterbit, Talend and Snaplogic will put more control in the hands of the business user and empower them to do useful things with the data from these disparate apps. Laypeople will be able to snap together data streams like lego blocks and make the things they need without involving a developer. The intuitiveness of the IDE for the lego-building apps will be paramount and a superior UI will emerge and become THE way it’s done (making one of those ETL companies a boatload of money). The other piece of the puzzle will be the presentation layer for consuming the data from these ETL apps. You’ll see more press releases like this one in which the presentation/collaboration product companies join forces with the ETL companies under the realization that peanut butter and chocolate just taste better together.
  5. Minority/Majority shift between desktop apps and web apps: I don’t have the current figures on desktop vs. web application usage (and I’m too lazy to look them up) but we’ll see a majority of one’s work conducted via the browser. This has been a trend in progress for some time but 2010 is the year that the perfect storm occurs where: connectivity improves sufficiently such that latency is negligible, web apps interfaces match the usability of desktop apps, there becomes a critical mass web-based alternatives for all former desktop-only apps and the ubiquity of access becomes crucial as necessitated by remote workers and telecommuting requirements.
  6. Trials become the new black: The traditional practice for ISV’s promoting a white paper that then promotes the download of their software will be replaced by landing pages that offer immediate trials right in the browser. The advent of mechanisms for delivering a fast & convenient hands-on experience will remove friction from the sales process. There will no longer be that step where the vendor needs to convince prospective users to expend energy to download & install software for the purpose of investigation.
  7. Social networking fatigue sets in and blogging sees a resurgence: People will get burnt out on the barrage of micro-updates from services like Facebook and Twitter and divert their precious thought cycles to fewer sources that serve as “lenses” and provide more depth. Twitter and FB will continue to experience insane growth and conversations will still occur via those channels but people will feel their mojo zapped and rediscover the .
  8. A major privacy breech casts doubt over enterprise use of SaaS for critical data: Cybercriminals will become more advanced and we’ll see a major breach of a high-profile SaaS provider like Salesforce. This will create a backlash that staunches the migration of IT operations to SaaS providers. The press will scream that the sky is falling, middle managers in IT will read articles and regurgitate headlines to CIO’s who will look for alternatives that deliver the same convenience factor of SaaS whilst satisfying the need to run on-premise. And JumpBox will be there to deliver ;-)
  9. Open Source gains mainstream acceptance: The stereotype of crappy UI’s and hard-to-use software will be gradually shed as apps like WordPress continue to deliver kickass user experience and win a huge number fans. Proprietary app vendors will cry, spread FUD and cling to a receding coastline only to see it inexorably washed away by OSS. There will still be a place for proprietary apps around niche situations but one by one the OSS substitutes for things like CMS’s and ERP systems will overpower their proprietary counterparts.
  10. An as-of-yet-to-be-discovered use of mobile phones becomes huge: In the mobile space companies will continue to build stuff nobody really wants (ie. ways to get spammed with location-specific coupons as you walk by a Starbucks). Meanwhile in a basement somewhere a small team will conceive and develop a killerapp for mobile that’s actually useful (either a consumer-facing app or a data mining app that’s sold to service providers). In the consumer space perhaps it’s a convenient 3-factor security mechanism that ensures your laptop can only be accessed when your bluetooth phone is with a few feet? Or maybe a clever way to facilitate ad hoc carpools amongst participants? On the data analysis side it may be a way for the CDC to model the spread of an epidemic via cell phones or a service for municipalities to do more intelligent traffic routing based on cell activity.

Do you agree or disagree with any of these? Do you have any predictions of your own you can share?
If you want more to ponder Read Write Web has some insightful predictions from its contributors. Here’s to computing awesomeness in 2010!

Tagged with:
Nov 17

So $3MM worth of iPhones were just stolen from a Belgian warehouse. The burglars apparently dropped in through a hole in the roof that was cut directly above where the goods were stored. They succeeded in getting away with the merchandise but given the precision of the location of the hole, it almost certainly narrows the list of possible suspects to those who had inside info on where the phones were stored. What should the robbers have done differently?

Answer: Cut holes in other places of the roof and tamper with windows and doors in surrounding areas to seed misleading evidence indicating that they were outsiders who cased the warehouse before discovering the iPhones.

Granted, they may have been pressed for time in their escape but by failing to apply disinformation they’ve decreased the pool of suspects and therefore increased the likelihood that they’ll be caught. So what relevance does this have to business?

The business case for disinformation

magicianHatIf you’re in a highly competitive space and you know you have competitors monitoring your activities, you’re likely making maneuvers that inadvertently telegraph your intentions. Public activities like domain registrations, trademark applications, patent filings and job postings can be spliced together to produce a picture of what you’re up to. The obvious recommendation is to conceal what can be concealed. But for those things which simply can’t be concealed due to their nature you can at least apply some creative slight of hand to obscure things.

Apple supposedly used a tactic dubbed the “canary trap” back in ’07 in which they selectively leaked false information via various channels to discover the internal mole that was the source for one of those Mac rumor blogs. There are digital rights management systems that use synonym substitution to create unique, slightly-altered versions of content. When false rumors are leaked via these documents they can be traced back to the source. Companies that have a crucial patent filing will often bury it in a haystack of red herring filings to obscure the move. And of course who can forget the famous heist sequence from Thomas Crown Affair in which individuals wearing identical outfits criss-crossed throughout the New York museum overwhelming authorities with suspects and eluding capture.

One of my favorite examples of disinformation was from Neal Stephenson’s book “Cryptonomicon.” The book covers – among other things – the story of how the German Enigma Code was cracked in WWI. Once the Allies had the ability to decipher Axis transmissions, a good deal of energy was expended responding to the intercepts in a way that concealed the fact they had actually cracked the code. They would have to stage a plausible scenario in which a Allied ship or plane would “stumble upon” a German U-boat that was discovered via a transmission. There was also a ploy in which they planted false information on the body of a deceased Allied officer and strategically placed it so that it washed up on the shore of the enemy to be discovered and assumed legitimate.

Other war time examples of disinformation are the numerous deceptions of the British officer Jasper Maskelyne. He made jeeps look like tanks, created the illusion of a battleship on the Thames and cloaked the entire city of Alexandria, Egypt from German bombers by building a small scale replica nearby, cutting the power to the real city at night and illuminating the replica. He would then dig fake craters and paint fake building damage in the night and to give German reconnaissance false assurance their attacks had succeeded.

The point of all this is that we are often so focused on improving the clarity of our message for potential customers that we neglect to take simple steps to obscure our movements from competitors.

What are some more examples of well-executed disinformation campaigns?

Nov 04

I periodically do video tutorials for the various applications that we package at JumpBox. The idea is to not just make the software easier to work with, but also to provide the instruction and motivation to help you get over the hump of doing something productive with it. With the one I did this morning we just hit the two-dozen mark and I felt like it was an appropriate milestone to do a “table of contents” post. In no particular order here are the videos (hover over the graphic to see the title and classification):

We have a nifty new feature that allows you to work alongside the tutorial by launching an instance on demand using only your browser. There’s nothing to download or install and you pay only pennies per hour for the time you use it. To learn more about that service go here. And if you find the videos useful and want to be updated as new ones come out, subscribe to our blog and tell a friend.

Oct 14

Here’s an interesting debate we had this morning in our office:

Would you consider this Twitter account SPAM?

Or the deeper question here: how do you define SPAM?

  • By a certain practice used to reach people?
  • By any unsolicited message with commercial-serving intent?
  • By a shotgun-style approach in communication?
  • By the relevancy of the message to the recipient?
  • It can’t be left to a completely relativistic definition because it becomes impossible to make laws to protect against it (ie. the one guy that happened to be wanting to buy viagra this morning finds the SPAM email to be very timely and useful, but that doesn’t justify the annoyance for the rest of us). On the other side of the continuum, it can’t be boiled down to specific practices because that’s what Bruce Schneier would call “the futility of defending the targets.” Here’s my position on the matter:

    I monitor key phrases on Twitter, certain sequences of words that indicate a user has a problem that one of our free JumpBoxes could solve. I skim hundreds of these tweets and select the few that we can help and respond to them individually introducing them to our product. I documented this technique awhile back. I’d say all but two of the 68 responses I’ve gotten from reaching out to people in this way have been received with appreciation. Two people have responded calling foul.

    According to the Twitter TOS the account above clearly violates the “If your updates consist mainly of links, and not personal updates” rule. But that could be satisfied by peppering it with personal updates and fluff. The reason I don’t do this out of my personal account or our JumpBox account is because doing so would inundate the followers of those with a bunch of repetitive info that’s uninteresting to them. But I digress. The point is there are ways to satisfy the TOS requirements but that just feels shady. I can see someone making the argument that this technique is not the “personal updates” spirit of use of what Twitter intended. I get that.

    But here’s what I don’t understand:

    • Making a freeware product recommendation for someone else’s product on a mailing list in response to a need that a participant expresses. Completely 100% kosher and expected.
    • Making a freeware product recommendation that’s your own on a mailing list when appropriate… cheesy maybe but still completely appropriate.
    • Making a freeware product recommendation of your own product in a distributed micro-blogging environment like Twitter where you single out a recipient who expresses a need your free product solves and you direct a thoughtful reply to that person… sorry but I see that as a legitimate way of reaching out to people. It’s not like you’re cluttering their inbox- it’s a message that appears on their @replies page in Twitter.

    If you were tying to sell them something- okay, I agree. If you were repeatedly harassing the same person- gotcha. But a one-time message that makes them aware of a solution that’s free and completely unique such that they would never know to search for it in the first place, I don’t see the SPAMiness in that. Anyways I’m probably going to be discontinuing this practice not because I think it’s spammy but because the return isn’t there time-wise.

    What do you think about this practice and the bigger question of how do we define what constitutes SPAM in the evolving world of social media?

    Oct 03

    preparation
    I have no idea if this story is true or not but it’s a neat parable on the value of preparation:

    A large top-tier law firm in New York was hiring a new attorney. They had taken hundreds of applications from recent law school graduates and had narrowed the search to three candidates, all of whom had first-rate GPA’s, achievements and LSAT’s. The firm flew the three candidates out over the weekend for an interview during which they were to present a mock brief to all the partners. The presentations would be the deciding factor of which candidate was selected.

    The candidates arrived on Friday evening and were given a tour of the office. The next morning they returned and gathered in the conference room where all the firm’s partners had assembled to hear the presentations. One by one they gave their 30-minute talks. Each one had done thorough research on the subject matter and had prepared compelling powerpoint slides. The first two candidates demonstrated supreme lawyering skills and “Perry Mason-like” courtroom demeanor. When the third candidate took the podium and it was immediately clear that he lacked the charisma of the other two.

    About halfway through his talk a gunshot rang out interrupting his presentation. Turns out it was actually the projector bulb on the conference table that had exploded. Given that it was the weekend there were no maintenance people on duty to replace the bulb. It seemed he would have to continue his presentation without slides. At this point however the candidate did something interesting – he calmly opened his briefcase and withdrew a spare projector bulb of the correct size and wattage. Within minutes he had replaced it and resumed his talk with his slides.

    Apparently during the tour the previous evening he had surveyed the conference room, noted the projector model and gone out that night and purchased a spare bulb as a contingency plan.

    All three presentations demonstrated thorough preparedness and while the third candidate lacked the superior speaking skills, his “meta-preparedness” sold the partners that he was someone who covered every base. The following week the third candidate received an offer to join the firm.

    preload preload preload