Bits of Kelly

So you’ve just received a Letter of Intent for a company to acquire your startup or for VCs to invest a round of capital into it. Congrats. But before you get to deal close – 30 to 45 days if you’re lucky – you have to slog through an intense period known as due diligence, where the acquiring (or investing) company will examine, probe, and question every part of your business.

I’ve been the diligence lead on two deals (selling SiteAdvisor to McAfee and more recently selling Hunch to eBay) and have learned a ton in the process.  I’d say that in both cases our business procedures and document organization were good – probably even very good – but they weren’t perfect, and there are things we could have done better.

I’ll outline a few tips here about what to expect and how to prepare for diligence, and how to successfully get through the process once it starts.

What to expect

* The first time you see a diligence request your jaw will probably drop. The ones we’ve received have been in the range of 400-500 initial line-item requests, with dozens more as follow-ups. A single line item could be as simple as a yes/no answer (“Is your company currently involved in any litigation?”) to something requiring hundreds of pages (“Provide all patent filings”, “Provide every employee invention agreement”, “Upload every material vendor contract currently in force or recently completed.”) In our recent deal we provided about 800 documents comprising 9,000 pages in response to diligence requests.

* You will be asked to put supporting statements and documents in a secure “Data Room”.  User access is controlled and you can set individual permissions for different folders. For example, you can grant access to salary or options-related information to only certain people with a need to know.  I used Merrill Datasite (no connection to Merrill Lynch) on both deals and am generally happy with them. Only gripe is that Mac compatibility could be better.

What you can do to be ready for due diligence:

* Use a well-known, big name law firm to provide you with a standard set of company formation documents, employee agreements, stock plan agreements, etc. We use Gunderson Dettmer. The key is to negotiate a low initial package price for these stock agreements (somewhere between $5,000 to $10,000) with the gentleman’s agreement that the firm will make the real money during subsequent financing rounds or an ultimate acquisition. As trying as the diligence process is in the best of cases, it will be exponentially worse if the other party’s counsel is questioning every core document because it was prepared by some unknown law firm or individual.

* Keep meticulous employee records for offer letters, stock option agreements, and especially PIIA (Proprietary Information and Invention Assignment) agreements. The PIIA in particular applies to EVERY current and past employee, contractor, or intern. The best control procedure for this is probably to ask your finance or accounting team to require a full employee checklist/document list before anyone can be brought on board or paid.

* Ideally you should have those docs already scanned and electronically organized. Hardcopies should be in a fireproof filing cabinet, with a copy given to your law firm.

* Ensure that all your documents and contracts are:

1) dated and signed by the other party, with a place for printed name as well.  (you’d be surprised how often a document gets misfiled and it’s impossible to decipher the signature to reveal who it belongs to)

2) dated and signed by your own company’s representative (this is one of the things we often forgot, although fortunately it’s an easy fix)

3) when it comes to responding to diligence requests, contract signature pages aren’t enough. At a minimum, the entire agreement needs to be included when requested. Ideally, each page before the signature page would also be initialed by the outside party.

* Keep all your vendor contracts in one place. Again, try to have them already digitized. Periodically ask around your company if anyone has started new vendor arrangements and not provided contractual paperwork to whomever is the central manager of that.

* Be mindful of formal vendor agreements that you let lapse, yet then continue under their current terms, thinking “let’s just keep going as we have been.” You’ll be asked for current, properly-executed contracts for all your current vendors. Try your best to always keep them up to date, formally extending them as needed, so you’re not caught scrambling for paperwork while the clock is ticking.

Getting through the process

* Assign one primary administrative executive who will lead the due diligence process, respond to requests from counsel, and coordinate resources to get everything answered quickly and accurately. You should generally pick an exec in your company who thrives on organization, has a substantive history with the company, and who has a good overall perspective about multiple functional areas. This process requires meticulous detail, patience, and follow-through.

* I find it helpful to create a Google spreadsheet with all the diligence requests, and then categorize them in one of 3 ways:

1) A straightforward question which can be answered with a simple statement or one sentence explanation. This can be answered right on the Google doc, a copy of which you’ll upload to the Data Room from time to time.

2) A request which is primarily satisfied by uploading existing documents.

3) A request which requires more substantive explanation and the creation of new documents/content.

-> Also assign each requested item to one core person in the organization who is responsible for completing it.

* There will be cases where essentially the same information is requested by different functions or departments. For example, HR and legal will have separate line item requests for employee/consultant invention agreements. But never duplicate uploads; provide the requested information in only one section and then refer the other request to that section. This both saves you money (you pay the Datasite based on every page you upload) and it means you need only update one area if you find a new document or need to correct something.

* When responding to requests – particularly requests that require the creation of new documents and content – think “explanation”, not “dissertation”. You want to address everything  with adequate thoroughness but not go into exhaustive detail. The idea is to disclose and explain, thus shifting responsibility for the knowledge of the item to the other party.

* Once most requests have been submitted and the other company has had some time to review, there will generally be a period of in-depth phone calls or in-person meetings to review follow-up questions. I find it helpful before each call to see which documents the interviewer has accessed from the Dataroom, so I get a sense of the starting knowledge the person likely has. The Datarooms have functionality to generate reports with this info.

* If you think nobody is really reviewing what you submit, think again. Counsel from the other company will be thoroughly reviewing absolutely everything to ensure you’re in compliance with their requests. I can’t tell you how many times I received an email at midnight saying something like: “Page 3 of 2007 intern John Doe’s employment agreement is not readable; please resubmit in full.”

Intense (and sometimes exhausting) as it is, the diligence process is also an incredible learning experience. It causes you to review everything about your company’s history and get a bird’s eye perspective of every department and function. And hopefully, everything will go smoothly and you’ll end up with an on-time and successful deal closing.

Here’s a newsflash which most startup marketers learn pretty quickly: nobody really cares that you just released feature set or design tweak #372 (unless you’re Facebook, in which case every tiny change you make will be over-analyzed with angry skepticism.)

If you want to generate some continuous buzz about your startup and its message, eschew incessant feature news and chest-thumping. In their place, try what I’ll call the ‘data flank’ approach: find a way to use your company’s proprietary, anonymized data to create a steady drumbeat of interesting insights.

Getting buzz from data flanks isn’t just about gaining immediate site traffic, sign-ups, conversions, or even revenue. All those things are nice and tend to happen to various degrees, but the real benefits are often longer-term and longer lasting. Your site/brand begins to be perceived as an authority with credible data. Consumers/end-users are willing to trust you. Potential partners learn about you and become amenable to business development discussions.

I’ll give you 3 examples below (2 from my own experience) how this can be done, followed by some concrete suggestions for how to do it well.

1)  A crowd favorite (and one of my gold standards for these types of data-driven insights) is online dating site okcupid, whose blog oktrends presses all the right buttons. It’s well-written, funny, based on rigorous data and analysis, and relevant to many of us.

This is a rather unique set of sexy data, because the data is for the most part about, well…sex. Still, reading okcupid’s excellent blog should give you ideas about how data can be presented in a way that’s entertaining, insightful, and at the same time reinforcing of the site’s core message (in this case, a big site used by interesting people who share honest feedback about their dating preferences).

OKCupid reports having 7 million active members and a whopping 1 million unique readers of its blog every month, an amazing statistic. The number of comments alone on each of its recent posts is testament to the power they have to engage readers, as you can see from some of these stats:

2) At my last startup, SiteAdvisor (later sold to McAfee), one of our marketing challenges was to take a relatively dry and obscure topic — web security — and try to broaden its appeal beyond just techy press and security forums. That meant trying to find a way to make web security fun, topical, relevant, and interesting to mainstream audiences with varied interests. We had lots of proprietary data about the safety of sites; we just needed to figure out how to make it interesting.

One of our early initiatives was to co-author an academically-rigorous report linking web security with a behavior that any internet user generally performs many times a day: search. Google “search engine safety“, and the report remains the #1 result even 5 years later. The study was featured in the Wall Street Journal, including a feature teaser above the fold on the front page of the print version. It helped us to both define, and own, the concept of “Safe Search”, which was the killer functionality of our software.

Having anchored our core message, we lightened things up a bit to broaden the audience. We began rating the safety of screensavers, grouped into topics that would resonate with different audiences.

We launched the World Cup of Spyware, pitting World Cup teams against each other in terms of which of their screensavers were more dangerous. That was picked up by sports writers. We did the same thing for MLB teams, American Idol finalists (pickup: celebrity/entertainment press), and a dozen other topics. A report called Mapping the Mal Web outlined dangers geographically, based on country-level domains.

Here’s an old press page showing some of our coverage in 2005 and 2006. Most of the links are now dead, but this still gives you a good sense of what we were able to accomplish. This page retains live links to some of the major studies.

The long-term results? A steady stream of broad-based press (which among other things helped to fuel the initial acquisition interest for the company), more than 100 million downloads of the SiteAdvisor app, and a subsequent major partnership with Yahoo.

3) At Hunch, we took a page from the SiteAdvisor playbook and continued producing reports and data-driven blog posts about all kinds of topics based on the proprietary data from Hunch’s Taste Graph. More recently, we’ve had the reports turned into professional infographics (which seem to be all the rage these days) using a professional outside agency.

The combination of interesting topics, solid data, and professional design has been a hit. For example, our Mac vs PC infographic was picked up by dozens of top-tier press outlets worldwide, and was on the homepage of CNN.com for 5 days, generating hundreds of thousands of visits to our site. It also helped generate inbound business development inquiries. Retweets referencing our reports continue for months after their initial publication, and popular reports have hundreds of comments from passionate readers.

Another example of results: here’s some of the press coverage for our interactive report on what your email domain says about you (click to see a larger version):

Ready to get started on some data flanking to market your startup? Here are a few practical tips to keep in mind:

1) Make an initial time investment to build some internal query tools so that your marketing or analytics team can explore data without involving the development team every time in one-off queries. Another approach is to make sure you have some of your marketers trained in basic data mining techniques, such as the ability to run straightforward sql queries on your database. (caveat: make sure there’s nothing they can do to crash your database; working on non-live data is probably safest)

2) Establish an editorial calendar for data-driven reports and insights and keep the ideas flowing with a steady drumbeat. (Cue the sports analogies like multiple shots at goal or times at bat.) Strive for a mix of evergreen topics and more topical themes that are pegged to news and major events.

Your best reports may very well go viral, but be careful about trying to predict which ones will. Instead, be ready for the winners when they emerge, so you can add fuel to the fire through levers you can directly control like promoted Tweets, Stumble Upon paid discovery, or BuzzFeed promotion.

3) Dial back the self-serving branding on most of your pieces and allow the data to take the spotlight. Think soft-sell. Notice how most of Hunch’s infographics have just small copy about Hunch and a small logo on the bottom right.

There’s an inverse correlation between the likelihood that media/press is going to pick something up and the square inches you take up with your branding and hard-sell messaging. Don’t worry, your attribution will come via links back to you, social media recognition, comments to your posts, and long-term business development benefits.

4) There’s an opposite corollary to #3. Rather than water down the effectiveness of all your reports/insights with drippy branding, choose just a few “internally-focused” pieces and go all out with your self-promotion. These should be the rare instances of chest-thumping at its finest.

Here’s how Foursquare did it to celebrate their impressive end-of-year 2010 metrics. At Hunch, we created a message-defining infographic explaining what goes into our Taste Graph. We didn’t expect it to receive tons of press (although it did get some), but rather it serves as a great “anchor the message” explanation for what we do, and we link to it all the time.

5) Pitch one outlet at a time. I don’t believe in spamming bulk lists of reporters or publications with every announcement or report we generate. We come up with custom lists every time, and generally write a personalized email to each person. This has helped us to build first name relationships with many great publications, bloggers, and reporters. Be especially careful if you use an outside agency that they don’t blast a pitch in a way that turns people off and is considered spam.

6) Packaging matters. Invest in great design to illustrate your data either through in-house talent or a qualified outside agency. You can get additional inspiration on sites like visual.ly, good.is, Many Eyes, and any number of infographic collection sites.

The examples above are primarily based on static, non-interactive data features. In a later post I’ll discuss some nice examples of interactive features that accomplish the same goals.

I mentioned to a friend recently that Hunch had just completed a very successful round of recruitment for summer interns, having attracted the most qualified candidates we had ever had.

Coincidentally, my friend had also recently heard from a client (CEO of a successful and growing software company) about what a hard time the client was having with recruiting in general.  So he asked me to take a few minutes to speak with the client to review her website and suggest some concrete recruiting tips.

These may seem really obvious to urban-based startups with a social media presence, but they were an eye-opener for this particular firm.  Among my tips to her:

• Recruiting shouldn’t be thought of as a single-shot approach for individual candidates, but rather as a brand-building exercise 365 days/year to accurately portray your company as a fun, energetic, plugged-in, rewarding, and challenging place to be.  Some of the ways we do that at Hunch:

1. We keep a Facebook page and highlight pictures of our loft office. We have an active blog, a Formspring presence, we showcase videos of the fun that our team has and we provide testimonials from past interns.
2. We actively use Twitter to reach out to our users and position our company as fun, edgy, and tackling challenging problems.
3. We always have a few extra desks, and we let entrepreneurs/friends use them while they’re getting started.  This increases the energy in the office with fresh faces, gives us smart people to exchange ideas with, and it gets people buzzing about our company.
4. We host MeetUp events at our office in the evenings to talk about geeky startup stuff, which attracts lots of interesting people, teaches us a lot, and at the same time gets the word out about us.
5. We host occasional Friday lunches for students (in fact, 10 Penn students came in today) where we talk about what it’s like to be in a startup and have informal networking with our employees.
6. We go to tech hack-a-thons to create cool new mash-ups and network with other hackers.
7. It’s not just about online recruiting. One Friday afternoon we sent several current interns to canvas some downtown Manhattan colleges and put up “old school” (paper) recruitment flyers in dorms and on common area bulletin boards. It worked: we received several strong applications and found a great new intern that way.
8. We find out who the student leaders are of college tech clubs, interact with them, and sponsor their events. They are motivated, incredibly helpful, and provide a direct link to talent without the filter of a campus career center.

• Although the client identified recruiting as her key short-term challenge, her website didn’t have a ‘careers’ link on her homepage, but rather buried in a ‘company’ sub-page. Easy fix.

• This client’s careers page had a very general statement saying “We’re always looking for people in the following disciplines…please send us your resume if you’re interested.”  This type of open-ended, wishy-washy statement is a turn-off to recruits. Is a particular position even open?  Is it worth a recruit’s time to research the company and craft a careful application? Any open position deserves an individually-crafted, stand-alone page which describes the position and its requirements in an interesting way.
• New recruits (particularly fresh out of college, starting their career), want to get a sense of the personality of the company and the people they will work with. But this person’s website only listed long, academic-like bios for two senior “management team” execs.  This might be reassuring for customers, but it doesn’t help with recruiting. Consider a simple team page which shows the breadth of experience and the general smarts of your team.
• We don’t have the administrative bandwidth to acknowledge and respond to every candidate submission.  But the second we interact with someone via a returned email, we designate that candidate as ‘active’ on an internal tracking document. We then track the number of days since their initial resume submission and since our last interaction with them. They don’t become ‘inactive’ until they either receive an offer or we promptly let them know that unfortunately we need to stop the process.  We never leave people indefinitely hanging after engaging with them. This is common courtesy but too often not followed.
• We don’t believe in free labor, although many candidates offer it.  Questionable legality aside, our philosophy is that quality candidates are expected to contribute in a meaningful way.  In return, we invest time and training, and we believe they deserve to be paid a competitive wage for their contributions.

Recruiting should be a continuous and long-term priority, but done correctly it should also be one of the highest return investments you make.

You’d be hard pressed to find a bigger skeptic about AOL than me, and my skepticism comes from many angles: as a former subscriber, a former advertiser, and even a former employee of the company for 2 years.

Yet with today’s announcement that AOL is buying the Huffington Post, and following AOL’s acquisition last September of TechCrunch, I’m wondering if maybe, just maybe, this dinosaur of an irrelevant brand might be able to be resurrected under Tim Armstrong’s watch. He certainly has kicked butt compared to what Carol Bartz has hasn’t done at Yahoo.

But I’m just not sure.  Despite making some bold and promising moves, Tim faces an uphill battle.  The reason is that the AOL brand itself has such a legacy of decline, mismanagement, and irrelevancy associated with it by many different constituencies, that I just don’t know if it’s salvageable. My advice: ditch it completely.

As I started this post I started remembering a bunch of specifics about AOL: the insane examples of dysfunctional management which occurred while I was a mid-level manager there, the wrath bestowed on the company by agencies and advertising clients who felt ripped off, the frustration of consumers as they tried to close their accounts, and the irrelevancy of the brand as millions of subscribers every quarter ditched AOL’s dial-up service to switch to broadband provided by someone else.

So much content came to mind that I though, hey – how about a series of posts on the subject?

Yeah, it’s easy to kick this maimed horse, but my real point will be to convey just how deeply damaged the core equity of the brand must be.  If I were in Tim’s shoes (and I am most thankful that I am not) the AOL brand would be out the window.

Thus marks the intro to my new, happy-go-lucky series entitled: “You’ve Got Stale: Why the AOL brand should be thrown out like last week’s curdled milk.” (I know, right? Catchy AND short.)

I know you’re riveted. Stay tuned.

We’ve been producing lots of interesting reports lately using Hunch data. Among them:

Voting in an Echo Chamber: The Myth of the “Informed” Vote

Aisle vs Window: The great debate

There’s also Magazines and Music, The Science of Attraction, and many more in the full report list.

Enjoy!

It’s confirmed: there will not be a Verizon iPhone available anytime before February 2011.

I know this because I had a 15 minute conversation with a local Verizon rep. while he was selling me a Droid X. I’ll call him Myron to protect my valuable source. This is serious intel.

"Ain't happenin'": A reliable source

“It’s happening, but not anytime soon,” Myron said, “And certainly not in January,” he added for dramatic emphasis.

How was he so sure?  He cited two reasons.  First, he says that secrets aren’t kept very well at Verizon, and if the phone were truly coming out in January, this would be the best-kept secret the company had ever had.

Secondly, he said that typically, training for major new products begins months in advance of their release, since customers start coming in and calling to inquire about the upcoming change. But there hasn’t been a peep about iPhone training.

These sound like perfectly reasonable conclusions to me.

So I’m discounting the fact that reputable media sources have followed the CDMA chip supply chain to note huge increases in core component sourcing.  I’m also ignoring the reports that contracted call centers have been asked to massively ramp up in anticipation of a major launch. And analyst speculation? Poppycock.

Could this conclusion have anything to do with justifying a decision to buy a pricey Droid X (with not much of a discount), because I just couldn’t stomach my POS LG “touchscreen” for one more day?  Could it have anything to do with the fact that Myron wanted to close a $350 deal today rather than see a customer walk until January? Nah.

So, draw your own conclusions based on “facts” and “senior management insiders”, but on this one, I’m going with Myron.

Since this was our first weekend in the city after closing up the Fire Island house, we decided to break out of our routine and venture uptown.  It was great.

We had dinner Saturday at Robert, in the Museum of Arts & Design on Columbus Circle.  (Warning: the restaurant’s website is really irritating.)  Food was great, interior design is terrific, views are amazing, and service was spotty but generally ok. We’ll definitely go back.

Our table at the restaurant. Amazing views of Columbus Circle, the Upper West Side, and Central Park.

The view of Columbus Circle from our table at Robert

After dinner we took a 10 min tour through the Hudson Hotel, which has an amazing outdoor terrace/bar.

Then we walked another 10 minutes or so to the new Ink 48 hotel on 48th & 11th.  Views from the 16th floor rooftop bar were absolutely stunning; there are no other tall buildings around, and it provided a rare East-facing view of Times Square and midtown.

Big City Lights

270 degree panorama shot taken with the help of my new Droid X

Nifty water feature has changing light colors

Sunday we headed back to the Upper West Side, starting out with a stroll around the Natural History Museum.

Museum of Natural History

Heed that warning

We passed by several nice-looking brunch places, but most were pretty dog-unfriendly and wouldn’t allow Nanu to sit under the table in their outdoor patios. Finally we found a place with less stringent rules. Our brunch at Al Dente was actually quite good, although the online reviews I read today are pretty bad, so I guess I wouldn’t recommend it in general.

Our meal was good at Al Dente, but online reviews are pretty bad

Next, we checked out an outdoor flea market. It was pretty upscale and they had some interesting stuff.

Old compasses, magnifying glasses, telescopes and sextants

Still, there were no buyers

Finally, someone’s short, little legs were getting tired. (And Nanu seemed to be getting pooped, too.) So we headed back down to Chelsea, with one of us hanging out the taxi window to greet passers-by.

Short legs are getting tired by this point

City girl on her way home