Business development: the Goldilocks principle

Background: At Hunch, we switched our focus (“pivoted“) about 14 months ago from B2C to B2B. Over that time, we pitched over 500 potential partners, trying to get them to use and eventually pay for our recommendation services. This process had its ups and downs, but eventually ended well when – after 8 months of grueling diligence – eBay decided to acquire Hunch in what I expect will be a successful outcome for both companies. During this time, I got a crash course in B2B sales/business development. Here is the first in a series of blog posts based on what I learned.

Somewhat counterintuitively, the biggest problem we encountered when pitching Hunch technology to potential partners wasn’t that it wasn’t interesting or useful to them, but that it was so interesting and useful that they considered it “strategic” or “core” and thus felt they needed to own and not rent it. The situation reminded me of the “Goldilocks principle” sometimes referred to in scientific contexts:

The Goldilocks principle states that something must fall within certain margins, as opposed to reaching extremes. It is used, for example, in the Rare Earth hypothesis to state that a planet must neither be too far away from, nor too close to the sun to support life.

Basically, if your technology is “too hot” – or, in business-speak, “strategic” or “core” – then there are three likely outcomes:

1. The potential partner turns you down because they decide to build a similar product themselves. This happened to us a number of times. I think part of the reason was that there was a lot of market buzz around “big data” and machine learning which lead to the perception – rightly or wrongly – that those capabilities needed to be owned and not rented.

2. The potential partner says yes because your assets are so defensible they can’t replicate them. I’m sure Zynga considers the social graph strategic but at least for now they have no choice but to partner with Facebook to access it. It is very rare for startups to have this kind of leverage, but ones that do are extremely valuable.

3. The potential partner wants to own what you do, but thinks you have a sufficiently superior team and technology that acquiring you instead of replicating you makes more sense. This is only possible if the partner is large enough to acquire you and has a philosophy consistent with acquiring versus building everything in-house. (A common tech business term is “NIH” which stands for “Not Invented Here”. It refers to a set of companies that consider anything developed outside of their offices technologically inferior).

At the other extreme, if your technology is “too cold” – perceived as not useful by potential partners – you’re going to have a lot of frustrating meetings. In this case, it is probably wise to reconsider whether there is actually demand for your product.

To build a long-term sustainable business, the best place to be is “just right” – useful to lots of partners but not so strategic that they are unwilling to rent it. This is where I wanted Hunch to be but we never got there. Most companies I know use externally developed products (commercial or open source) for databases, web servers, web analytics, email delivery, payment processors, etc. These are often highly competitive markets but the companies that win in these markets tend to become large and independently sustainable. These “just right” companies – to extend the astronomy analogy – are the planets that support life.

Anonymous

Regarding point 1, that’s really just too bad for the partner. As any longtime Hunch user knows, the recommendations the engine produces are just so much better than almost every “built-in” recommendation system out there. I think maybe only Amazon comes close.

The fact that these capabilities can provide benefits if “owned” doesn’t necessarily mean that your company is capable of doing a very good job of it.

It all sounds like a really interesting case study on customer development to me – even if you can provide the “magical” solution, sometimes the customer still doesn’t want it!
Anonymous

Regarding point 1, that’s really just too bad for the partner. As any longtime Hunch user knows, the recommendations the engine produces are just so much better than almost every “built-in” recommendation system out there. I think maybe only Amazon comes close.

The fact that these capabilities can provide benefits if “owned” doesn’t necessarily mean that your company is capable of doing a very good job of it.

It all sounds like a really interesting case study on customer development to me – even if you can provide the “magical” solution, sometimes the customer still doesn’t want it!

http://wearenytech.com/64-mark-birch-investor-entrepreneur-trader Mark Birch

One of the things you are not considering is the internal politics involved in enterprise or B2B technology sales. There is a constant battle between different factions, whether it is business and IT, different lines of business, domestic and international, distributors and partners, etc. Existing relationships with other vendors at various levels of organizations can have a significant impact on your ability to get consensus needed to secure a deal, ESPECIALLY when the technology in question is considered strategic. I ran into this very issue in my enterprise sales experience and found that build vs. buy was generally the biggest competition, right next to “not a business priority”. This is why everyone you pitch to in a room will nod in agreement, tell you they love your technology, and say they will champion your cause, but months later you have gotten no closer to advancing the sale.

http://www.cdixon.org chris dixon

I agree with you and think I’m aware of that issue (and plan to blog on the topic in the future) although sounds like you have a lot more experience with it so would love to hear more of your thoughts.

Anonymous

I try not to miss an opportunity to refer to Marc Andreesen’s “Moby Dick theory of big companies” http://pmarca-archive.posterous.com/the-pmarca-guide-to-startups-part-5-the-moby

http://twitter.com/andyidsinga andyidsinga

great link Saul ..and so so so true of big companies and startups.

Chris, looking fwd to more posts on the subject.

http://twitter.com/cselland Chris Selland

So true – the need to work through those internal politics means everything. A small number of ‘no’ votes – sometimes just one and often for very parochial reasons – can easily kill a deal or at least postpone it indefinitely. NIH is frequently a factor – finding a pro-deal advocate who is influential on the product and/or tech team can help keep things moving – or at least give some insight into why they’re not.
http://www.repeatablesale.com/ Scott Barnett

You are so right, Mark. This is where (IMHO) many B2B entrepreneurs get tripped up. The best produce almost *never* wins. It’s very much a relationship sell, and figuring out how to navigate the internal corporate politics. That is why B2B sales take months instead of minutes/hours/days with B2C. More decision makers, many more moving parts, and you need to figure out who has influence, knowledge, power. In addition, what are the internal motivations of each stakeholder. As Chris said, these large partners have to be big enough to buy you, so it’s unlikely anybody at that company has the personal stake invested that you do. So, their motivation isn’t the same as yours…. it could be greed, fear, power, etc. That is where EMPATHY comes in… if you understand the needs and motivation of the person across the table (instead of what YOU think they should be thinking) then you have a better chance to win.

http://www.cdixon.org chris dixon

Thanks – that’s very nice of you. I always thought so too but obviously I am massively biased.
http://johnbpetersen.tumblr.com John Petersen

I completely agree Nate. I come from an international bank where the IT department had such pull over product decisions that they almost always determined that it was a better idea to “build” instead of buy or rent. They obviously had a very vested interest in keeping their people busy. Kinda sad actually.

Without fail, they would build an inferior product that was over budget and behind schedule.

I am very curious to hear if Chris has any insight on where some of the companies are that decided to build their own engines are now. I expect similar results.

In cases like my former company, you could be a perfect fit and even convince what seemed like the right people to go with you but once it got to certain decision-makers (the heads of IT in my case), you didn’t stand a chance.

http://twitter.com/kn_shiva Shiva

Waiting for the next in the series…I totally agree about dynamics of partnerships for such a product.

http://www.cdixon.org chris dixon

ok thanks i guess that’s positive feedback a bunch more posts coming but would love more commentary as to what is interesting to go deeper into.

A Close

Dunno if that’s an open invite, but if so, I’ll bite. ”We pitched over 500 potential partners”. Curious about that large number and how the pitch may have evolved, any specific lessons learned? 500 seems like a lot to me, was there a method to the first 100 and different tactics for the second 100, etc?
Congrats!

http://www.cdixon.org chris dixon

Definitely had a very systematic method. 500 might sound like a lot but over 18 months is a couple of conversations a day. Also, a lot of those were exploratory. Take daily deal companies for example. We talked to probably 10. We discovered pretty quickly that 1) geography was main predictor for venue-based daily deals (people respond to offers that are physically near them), 2) the big guys see recommendations as strategic (“too hot”), and 3) most of them (at least at the time) didn’t have enough offers per day per city to do a lot of interesting targeting beyond geography and maybe gender. So we pretty quickly decided it was a bad category for us.

A Close

Interesting. To have the big daily deal companies seeing recommendations as “too hot” seems quite sluggish IMHO. But maybe it’s a case of don’t fix what isn’t broken.

Thanks for your incites.

http://wearenytech.com/64-mark-birch-investor-entrepreneur-trader Mark Birch

“technology is “too cold” – perceived as not useful by potential partners”

While this can seem frustrating at first, sometimes these can be the best customers, as you are building the vision in their minds before they have developed their own thoughts or gotten infected with competitors offerings. In competitive markets, this is less likely, but in green field, nascent markets, it is easier and often times necessary to walk through the “education” process in order to secure sales.

At any rate, you are absolutely right, B2B technology sales is not something to be underestimated.

http://www.cdixon.org chris dixon

Yes, “too cold” can mean “just not ready”. You have to skate where the puck will be etc etc. a cliche but true.

http://twitter.com/lindadrabik Linda Drabik

May be worth exploring the pitfall of many B2B efforts: begins as just
right, but gets bumped from rentable to cold- not because it’s not useful, but
because stakeholders aren’t on the same page about requirements and ROI of
adoption. Goldilocks morphs into too many cooks in the kitchen.

http://twitter.com/ferodynamics PJ Brunet

Nice morphing metaphor.

http://twitter.com/lindadrabik Linda Drabik

Agree with your earlier comments re: build vs. buy and switching costs due to
incumbents.

For green field, the hurdle seems to occur in that space
between ‘how do we adopt/integrate’ (or in the case of moving enterprise to SaaS,
objection handling re cloud migration of sensitive data or fresh start) and truly turnkey.
Integration hopefully doesn’t get pulled down the slippery slope of other
initiatives, while truly turnkey often occurs once product/market develop and
other soln’s begin to enter or catch up.

To get to the ‘juuust right’ space early on, the technology
provider can take a modular approach to its solution, breaking out and pricing
the part of the offering that’s the true fit to simplify/curb the barriers to
adoption. That approach may limit scale since modular = customized before product proven and core defined, but protects against some of the issues above.

http://unystartups.com Julian Baldwin

Nice post Chris and congratulations on the deal with Ebay.

It seems to me that in the early days of certain new technologies it’s best for companies to compete to build and own the technology outright to promote innovation downstream. This might be perceived as good because it encourages the rapid development of a given technology before it encourages the rapid integration of it.

During the rapid development phase a technology can be discovered, learned about and tweaked for performance issues associated with scaling. When the rapid development phase precedes the rapid integration phase, future concepts for innovation have more stable ground to stand on and thus innovations can emerge more regularly.

In the rapid development phase a few entities wrestle to establish the standards and protocols of said new technology so others can then cut and paste bits and pieces of it into a testbed of applications. This I would associate with rapid integration as well as your reference to commercial and open-source software for databases, etc.

So as it relates to Hunch, my guess is the technology is still very much in the rapid development phase of the innovation pipeline. This would (to some degree) explain why you encountered the experiences you did when trying to obtain partners. My guess is ten years from now the rapid integration phase will be well underway.
David Whiteman

Chris, my experience with Hunch did not quite neatly fall into one of those categories. While we viewed your tech as strategic and important and signed an agreement with you, it never made it into any of our applications. What Hunch has built is unquestionably smart and interesting but I struggled to move it into our list of ‘core’ functionality. Being a resource-constrained business made it difficult to spend the necessary time to figure out how to best integrate your tech. Invariably, it would end up in the ‘too-hard’ bucket at planning time. You can argue that it was just not useful enough for our business to prioritize, which is partially true, but I also think your team may have underestimated the level of education and evangelizing required to drive adoption. I suspect part of that was to do with your own resource constraints and your focus.

That said, I’m disappointed I won’t get the chance to take Hunch for a test drive, unless of course eBay is willing to play nice and share? I think the acquisition is a brilliant move by them for the talent alone. Hope you guys get enough room to play and keep doing what you’re doing.

http://www.cdixon.org chris dixon

Thanks for the comment David. Yes, I agree these categories are overly clean. In reality a lot that we did wrong was probably due to human failure including especially my own. The good news is the plan is to keep the Hunch site and API open at least for the foreseeable future…

David Whiteman

Chris, my experience with Hunch did not quite neatly fall into one of those categories. While we viewed your tech as strategic and important and signed an agreement with you, it never made it into any of our applications. What Hunch has built is unquestionably smart and interesting but I struggled to move it into our list of ‘core’ functionality. Being a resource-constrained business made it difficult to spend the necessary time to figure out how to best integrate your tech. Invariably, it would end up in the ‘too-hard’ bucket at planning time. You can argue that it was just not useful enough for our business to prioritize, which is partially true, but I also think your team may have underestimated the level of education and evangelizing required to drive adoption. I suspect part of that was to do with your own resource constraints and your focus.

That said, I’m disappointed I won’t get the chance to take Hunch for a test drive, unless of course eBay is willing to play nice and share? I think the acquisition is a brilliant move by them for the talent alone. Hope you guys get enough room to play and keep doing what you’re doing.
David Whiteman

Chris, my experience with Hunch did not quite neatly fall into one of those categories. While we viewed your tech as strategic and important and signed an agreement with you, it never made it into any of our applications. What Hunch has built is unquestionably smart and interesting but I struggled to move it into our list of ‘core’ functionality. Being a resource-constrained business made it difficult to spend the necessary time to figure out how to best integrate your tech. Invariably, it would end up in the ‘too-hard’ bucket at planning time. You can argue that it was just not useful enough for our business to prioritize, which is partially true, but I also think your team may have underestimated the level of education and evangelizing required to drive adoption. I suspect part of that was to do with your own resource constraints and your focus.

That said, I’m disappointed I won’t get the chance to take Hunch for a test drive, unless of course eBay is willing to play nice and share? I think the acquisition is a brilliant move by them for the talent alone. Hope you guys get enough room to play and keep doing what you’re doing.
Pingback: Business development: the Goldilocks principle | | BUSINESS VOCABULARYBUSINESS VOCABULARY
http://metaoptimize.com Joseph Turian

This. NLP+ML are hot right now, because they are valuable if applied correctly, but this hotness makes it difficult to rent technology. I’ve experienced this consulting through MetaOptimize on NLP and ML, and it’s the biggest barrier to acquiring clients. I posted a more detailed commentary on my experiences on Hacker News: http://news.ycombinator.com/item?id=3288631

http://www.cdixon.org chris dixon

Thanks for the congrats! I think many more blog posts could be written on what’s going on in ML right now and how it’s trendiness has screwed up the market.

http://metaoptimize.com Joseph Turian

That’s a good suggestion. I was planning on amping up my blogging about technical how-tos and general trends and news, but the business side of things is a good topic too.
http://metaoptimize.com Joseph Turian

That’s a good suggestion. I was planning on amping up my blogging about technical how-tos and general trends and news, but the business side of things is a good topic too.

http://wearenytech.com/64-mark-birch-investor-entrepreneur-trader Mark Birch

By the way, it is great to see you blogging again. I always greatly enjoy your posts (even if I disagree on occasion).

http://www.cdixon.org chris dixon

Thanks Mark! If you didn’t disagree at least 1 in 5 times, I’d be doing my job wrong!

Irving Fain

Great post Chris! Nice to see you talking about the B2B and enterprise side of start-ups…

The other downside to being “too core or strategic” is the effect on the sales cycle. When something is perceived to be so valuable and/or so important to day-to-day business, stakeholders in the organization become paralyzed. They’re both enamored with the technology and its potential, but at the same time fearful of its influence and effect on what they’re doing and how they’re doing things already. This seems to be why there’s such an insurgence of technology companies (particularly in the social marketing space) offering singular and oftentimes narrow features. It’s both easier for people to get their head around as well as easier to make a quick decision. Of course this also ties into the build vs. buy argument that big co’s know and love…

In the enterprise / B2B space nothing is more valuable than perseverance. Happy to see such a great outcome for you and the team. Well deserved!

http://www.cdixon.org chris dixon

Thanks Iriving! Good points. One more reason why b2b kind of sucks.

Irving Fain

B2B is hard, but it’s still important. I think there’s too much focus right now on the seamless, frictionless and oftentimes B2C plays because those move up the adoption curve more quickly. That said, our economy is built on businesses selling to one another along with the consumer (look at the distributor model present today in retail, liquor, etc.). Somehow, in technology, this has definitely fallen out of favor since it’s slower moving than the rest of the tech world….

http://arnoldwaldstein.com awaldstein

Sometimes, being hot has little to do with the outcome.

I spent some time buying startups for a larger public company. The goal was to acquire pieces of a portfolio of solution for the enterprise that could be bolted together.

What I was buying was some market proof for a solution that my company’s bigger reach in sales and current customers could capitalize on.

The upside for the acquirer is that we paid less for the potential of a new revenue stream. The upside for the acquired company was that if right, the earn-outs could be substantial with the added resources.

Greatest challenge was culture clash not product or technology applicability. What made it work, when it did, was the acquirers sales skills to handle both the acquirer’s politics and the customer base.

Different from your example, but still relevant for startups to keep an eye towards.
Matt Greenberg

I’ve got to fundamentally disagree with you about “too hot” technologies and components. So many technologies now are deployed “in the cloud” where the whole infrastructure they are doing is owned by someone else and they are just “renting” it. Smart companies rent in perpetuity all the time.

http://www.cdixon.org chris dixon

Something can be hosted in the cloud (e.g. on AWS) and still owned by the company. I think you might be confusing legal/strategic ownership with physical proximity.

http://twitter.com/ferodynamics PJ Brunet

As long as AWS doesn’t lose their data

http://www.cdixon.org chris dixon

Something can be hosted in the cloud (e.g. on AWS) and still owned by the company. I think you might be confusing legal/strategic ownership with physical proximity.

http://twitter.com/leahculver Leah Culver

Thanks for the excellent blog post. This is actually something I’ve been thinking about recently but not quite able to put into words.

At least, I’ve been thinking a lot about the “too hot” side of things in relation to other companies trying to sell us their real-time web solutions (for Convore). Most of those definitely fall into the #1 category not because we felt we needed to theoretically own the technology but because it was much easier to build it quickly ourselves than to integrate it. Or does that mean that it’s “too cold”?

I’m really looking forward to reading more posts on your transition from B2C to B2B since I’m in a similar situation myself right now (Pownce/TypePad/Convore -> Grove.io).
http://www.facebook.com/dkhammer David Hammer

One ideal way to tackle the goldilocks problem is to find a company (or even better, an entire industry) that has meaningful technology requirements but has learned that it can’t build technology of the sophistication that the company can provide. Focusing on technology rather than data is maybe less defensible in trying to build a Facebook-sized business (eventually competing tech companies will see the opportunity and jump in), but I feel there are many large businesses that can succeed by catering to technology-backed industries that don’t think they have the organizational ability to build technology in-house.

http://www.cdixon.org chris dixon

I agree. I think this is the strategy of selling to “old school” businesses like insurance companies or consumer goods. The problem there is often the opposite – you could have invented the best thing since toothpaste and it takes them 3 years to realize it and by then you’ve run out of capital.

http://www.ipatient.com Dan Munro

This is exactly what happens in healthcare – sadly – a lot.

http://twitter.com/dbenyamin dbenyamin

I think you are missing a huge segment of case 1, namely that the technology can’t be adequately ‘ componentized’. Through experience at my own ML & NLP startup, there are many domain specific tweaks needed for anything considered high performance results. I shudder at off the shelf text analysis, since ‘text’ could be an IM, a tweet, a blog, or a novel…

Do you really think hunch worked equally well for those hundreds of beta users and exceeded their own enterprise knowledge?

http://www.cdixon.org chris dixon

That’s a complicated question. Without adequate input data (which partners were often reluctant to give), it was harder for our significantly better technology to beat their commodity technology since we had such a big data disadvantage. I found a lot of the resistance to be due to programmers who took a few statistics classes and convinced their bosses they knew about machine learning and then prevented us from ever getting the data we needed to prove how much better our algorithms were.

http://twitter.com/dbenyamin dbenyamin

Tough path, agreed!

http://twitter.com/vsagarv Vijaya Sagar V

Congratulations to all at Hunch, Chris. Glad to see you write again. ‘Was wondering whether your responsibilities at Hunch were weighing you down so much …

There’s another outcome to add to your list, a diabolic variant of #3. A large company buys off a smaller but ground-shifting technology company to thwart a shakeup in the status quo. The incumbent feels threatened when an “upstart” appears to rewrite the rules of the market. Ofcourse it means that the large company has weighed outcome#1 considerably before giving it up. It is also possible that the large company has some secret thing in the works already and this new competition puts a spanner in their work (a diabolic variant of #1).

And, I feel it is more fulfilling and clean to win the 2C market than 2B, because one does not need to ‘impress’ a few individuals whose intentions are not always transparent. No need to have unwanted dinners at strange places, when the evening is better spent over a bit of jazz, watching new user love collect in the inbox.
Anonymous

I did biz dev for algorithmic software for 10 years, so I can sympathize with the challenge. In my experience, if the technology is truly perceived as core, it needs to be crushingly superior for you to be able to overcome the political, emotional and rational hurdles.

If it’s not it’s likely a waste of time to try to push through a deal. If it is, it’s often a good idea (to avoid benchmarking hell) to negotiate an LOI before you start, covering:
- The shared data set to be used for benchmarking: yours, theirs or public domain
- Success metrics
- Next steps given success

If they’re not willing to do that you need to ask yourself whether your BD person is qualified.
- If yes: move on to a new customer (unless circumstances are somewhat unique)
- If no: upgrade BD person (finding a killer bd person may or may not be an option. There’s a reason why the really good ones usually have CEO on their business card – they are equal parts sales person, product person, marketer and engineer and people like that are rare)

Congrats on the deal.

http://www.cdixon.org chris dixon

Good advice – thanks.

Pingback: Learning From Others
http://twitter.com/baetsie Benjamin De Baets

Thanks for this honest look behind the scenes, though I’d really like to read an in-depth post about the decision to “pivot” from B2C to B2B.

In my opinion that’s the most important point in Hunch’s existence. Of course I’m hugely biased, but it makes me somewhat sad that the website never managed to gain significant traction. Imagine an exponential growth combined with your back-end technology, resulting in a taste graph covering all demographics.

But given the circumstances, this is probably the best possible scenario.

Btw all the visualizations, applications, cool demos that were built to convince possible partners, any chance Hunch users will ever catch a glimpse of these?
http://www.chriskurdziel.com/ Chris Kurdziel

Chris,

Congrats on the acquisition! Nice post and glad to see you blogging again. Particularly agree with Mark Birch’s thoughts here on the politics of the enterprise – pretty consistent with my experience as well.

Would be really curious to hear some more thoughts about the pivot to B2B and how it affected other parts of Hunch, particularly the culture of the team during the months/weeks of transition.
Anonymous

Thanks Chris–very insightful. Many corollaries between Hunch and 23andMe: a consumer-driven database of associations and connections that potentially holds high value to enterprise (in our case, pharma/biotech).

Spent most of my career in 2B, 2A (academic) and 2G (govt); leaving that world to start a 2C company was a breath of fresh air. But now 23andMe faces a similar challenge: how to leverage their valuable, unique database toward more personalized therapeutic design (without alienating their consumer base). Pharmas are highly protective and controlling of their drug IP; handing over the keys to a new development strategy based on an external partner’s resource may not fly. Time–and many protracted discussions–will tell.
http://rc3.org/ Rafe

The other problem I see a lot in B2B is that not only does the customer have to pay cash to use your product or service but that integrating with it is also a project that has to be sold to the IT guys. Even if the stakeholders on the business side think the value is there and the fit is perfect, the IT guys can torpedo the deal, often simply by giving an unreasonably high estimate for how long it will take to complete the integration. Having technical people involved in the sales process who can convince the IT people to get on board is huge.

http://www.cdixon.org chris dixon

Totally agree. I was thinking my next b2b post might be about the difficulties of integration.

Pingback: cdixon.org ? chris dixon's blog / Business development: the … « poborakinisi
Pingback: cdixon.org ? chris dixon's blog / Business development: the … « irocynanos
Pingback: Interesting Articles of the Week – 11/28/11 | Royce Hamano
Pingback: The evolution of search, Facebook’s IPO, and Bing’s “Panda”
Pingback: Trying to get to “just right” | Melih's Thoughts
http://twitter.com/marco_can Marco Cantarella

I agree with this as well. As someone who has come up through B2B business development initially I would waste countless hours chasing any client who I thought had the pockets to pay. However after defining the boundaries you come to “the Goldilocks Principle” which you have put so elegantly Chris.

In future posts I would love to see some info about how you came to that conclusion, some of the false positives that you found and are there exceptions?

Look forward to reading the series.
Pingback: Forces that affect whether a large company will buy your product (according to Marc Andreessen) - Chris Dixon
Pingback: My year in blogging - Chris Dixon
Pingback: Quora
Pingback: Takeaways: Alex Taub’s ‘Intro II Business Development & Partnership for Startups’ Skillshare class « Ent!
Pingback: 4 Machine Learning Sessions at Structure:Data that Shouldn’t Be Missed – MetaOptimize

Chris Dixon