Startup Marketing: Big Bang vs. Darwinian Evolution

About This Blog

This site is for  entrepreneurs.  A full RSS feed to the articles is available.  Please subscribe so we know you're out there.  If you need more convincing, learn more about the site.



And, you can find me on Google+

Connect on Twitter

Get Articles By Email

Your email:


Blog Navigator

Navigate By : 
[Article Index]

Questions about startups?

If you have questions about startups, you can find me and a bunch of other startup fanatics on the free Q&A website:

Subscribe to Updates


30,000+ subscribers can't all be wrong.  Subscribe to the RSS feed.

Follow me on LinkedIn


Current Articles | RSS Feed RSS Feed

Startup Marketing: Big Bang vs. Darwinian Evolution


Broadly speaking, I think there are two types of marketing approaches. 

The first is what I'll call "Big Bang".  This was very popular in the 1990s, particularly for venture-funded startups.  In this approach, the sequence of events goes something like this:

1.  Have idea
2.  Raise Capital
3.  Code like crazy in "super stealth mode"
4.  Hire VP of Marketing to plan big launch
5.  Hire PR agency to launch
6.  Launch!
7.  Success!(?)

I'm not a big fan of this approach for one simple reason:  I don't think it works all that well.  As a developer, I think this is a wee bit like trying to write a 200 page specification document, develop the product, and expect to release something that works and makes users happy.  It just doesn't work that way.

The second approach, the one I do like, is more like Darwinian evolution.  This is where you start as early as possible, experiment as much as possible, as efficiently as possible and respond to feedback as quickly as humanly possible.  Keep doing more of what works, and less of what doesn't.  In my experience, this works really well.

So, if I were working on a startup (which it turns out, I am), I'd lean towards a strategy that looks more like this:

1.  Have idea
2.  Bootstrap / Beg / Borrow
3.  Tell the World
4.  Release product to the unsuspecting
5.  Get feedback
6.  Iterate, iterate, iterate!
7.  Success!

I'd like to spend a little bit of time on the middle parts of the above sequence.  What I think really works today is discussing the idea with your potential market as early as possible.  Ideally, this should happen before you've written your first line of code.  Less ideally, you can wait a week or two.  The easiest way to do this is to start a blog.  You can use one of the free services out there (just make sure to register your own domain name).  This way you get commenting (market feedback) and RSS subscriptions automatically.  Then, post like crazy and do everything possible to get feedback. 

In parallel, put the most minimal version of the software imaginable out there as soon as you can.  If you're not embarrassed of your product and are not scared to death when users start banging on it, you waited too long.  Get it out there.  Yes, people are going to point and laugh at your product.  Yes, they're going to ridicule you for building something that has a laughably small set of features, most of which don't work.  It doesn't matter.  GET IT OUT THERE! 

From there, your marketing and your product should evolve in tandem -- organically.  Don't even think about advertising, PR, launches and other marketing stuff.  All of these things will simply distract you from the real problem:  figuring out what customers want.  As many failed startups have learned, marketing a product that nobody really wants is awfully expensive and frustrating.  Even the perfect marketing strategy (if there is such a thing) is unlikely to work.  On the other hand, even naive, unsophisticated marketing can work wonders when everybody wants the product.

So, don't get lured into believing you need some super-sophisticated marketing strategy with a big launch to create a mega-hit product.  You don't.  If you don't believe me, think of five startups that you really admire and that you think were big hits.  Now, do some quick research and figure out how much time/energy they spent on a big-bang launch to "release" their product to the world.  Chances are, the launch came well after the product was already out there and somewhat successful.

What do you think?  Have you mastered the art of launching a successful software product?  Are you a PR person that is gravely offended at my suggestion that startups shouldn't really plan formal launches?  Would love to read your thoughts in the comments.

Posted by Scott Stephenson on Mon, Jul 16, 2007


Your strategy is good, but the comparison of the "1990's" approach to your suggested approach indicates that step 7: success is a given. All business startups are a risky venture. The possibility of not being very successful at all is very high.

posted on Monday, July 16, 2007 at 8:32 AM by Steve McLeod

I agree the evolutionary approach works well for web-driven startups like the current crop of Web 2.0 companies. The software is fairly simple (developed in months not years) and the service can often be given away free or cheap in a beta stage.

The evo model doesn't work well at all for many other startups e.g. if you are developing enterprise software with long sales cycles and demanding customers, or a more complex product or service (e.g. a piece of hardware like the Slingbox, a service like Netflix or a business based around a licensing model like Dolby). A customer driven product is essential, but in many non-Web companies you simply cannot expect to sell a product that isn't ready for the market.

You may also find your VC's very uncomfortable with a less structured approach.

posted on Monday, July 16, 2007 at 11:24 AM by Todd

My background is in the world of business development and marketing. Basically grew a company called MindSpring, launched EarthLink's mobile wireless business, and helped grow a company called CipherTrust that was purchased by Secure Computing for $274 million. In three words "creep don't leap" is the best marketing approach. Make something somebody loves and they will tell their friends about it. Marketing is much more about product then advertising or pr.

posted on Monday, July 16, 2007 at 12:24 PM by

Dharmesh, I like all your points; especially #7 ;-) but my question, especially in my own context is: how can you best do #5? Get feedback implies comvincing a decent number of users to try out your site/app - how are you doing that? Starting a blog is an idea but how do you get folks to read *that*? Please don't think I am trying to be a wise guy - I'd really like to know.

posted on Monday, July 16, 2007 at 6:07 PM by Raghu Srinivasan

Raghu: This is a fair question. Tactically, what I've found to work is: 1. Contact folks you know would likely be interested (often these are people from a prior life that you know). 2. Leave comments on related blogs that contribute to the conversation (and have a link back to your content that is relevant) 3. Consider spending some some modest dollars on something like Google AdWords to get targeted traffic back to your site. Unfortunately, there is no easy answer to this. In my experience, as long as you are not making the "hard sell" and genuinely looking for advice/feedback, things go a little easier. People can smell a sales pitch from a mile away.

posted on Monday, July 16, 2007 at 7:29 PM by

Dharmesh - Completely agree with you here. Also, I think that in small web 2.0 startups, the pendulum has swung too far to the "we'll build what we need and not features based on customer input" side. If you are building in stealth mode, there is no way the features are based on user input.

With our new product, I'm taking the approach you recommend here - combined with a narrow customer target (sales team collaboration) - so that I can implement features important to the people that will be buying the app. I am blogging about it each step of the way and posted scope outlines and screen sketches before starting coding. I've found that the rapid, ongoing feedback cycle is a great thing.

posted on Monday, July 16, 2007 at 10:21 PM by Scott Meade

I plan on using the evolutionary approach. I just finished a FastTrac class. In my "investor presentation", I had "low" numbers for marketing because I plan on using the approach you describe above. Even though I plan to do B2B for financial services, I don't think I need to go the "traditional" route. Thanks for the info.

posted on Monday, July 16, 2007 at 11:12 PM by George

I am in agreement with the evolutionary approach. It's exactly what I am doing now. I am almost to #4 right now. I started to blog about my upcoming product about a month ago. We have a couple of features that are about to go public beta, and I will probably get laughed at at first. But the criticism will be good. Also, I am going to do this with no marketing; pure organic.

posted on Tuesday, July 17, 2007 at 1:04 AM by Jeff Rohrer

I totally agree, I released the first version of my current product after just two weeks of work, I've got excellent feedback from people using it and kept improving it based on user's requests.
I'm convinced that in most cases starting very small and continually improving is much better then the alternative.

posted on Tuesday, July 17, 2007 at 4:15 AM by NirD

I agree with the approach .. but just a point ........ #3 and #6 need a lot of discretion, especially, if you want to retain the first movers advantage. ofcourse it depends on the scenario.. we use a similar approach as mentioned by Dharmesh.. but careful discretion is something totally important through the process of iteration and bootstrapping in our case even though we are looking for feedback at every step. can be tricky.

posted on Tuesday, July 17, 2007 at 10:31 AM by

Like everyone else posting comments, I completely agree with going iteratively. I've always put an emphasis on trying to generate a mountain of buzz in the beginning, but I'm learning that it's not the most ideal. Especially when you don't have the cash to burn on a marketing plan.

Sometimes the feedback is tough to hear and sometimes you get annoyed with what people have to say, but in the end once you've sorted through and got to the valuable feedback you'll find some nuggets of information for you to work with to make a better product.

posted on Tuesday, July 17, 2007 at 6:06 PM by Tobin

As one of the founders of a bootstrapped startup, I would say you have just covered the launch of a company/product. In either of the cases there are probably other steps between 6 and 7.

posted on Wednesday, July 18, 2007 at 6:53 AM by Ashish

Very good post - and the second scenario is truly where 90% of tech entrepreneurs should devote their efforts. My question is what to do in regards to competition. Let's say you've already launched phase I of your product, listened to your customer, and found a few killer apps that will put you leaps and bounds above your competition. In this position, we've opted for total secrecy until it launches (small lead time of 6-8 weeks). We figure it's best to totally catch the competition off gaurd, and bloggers will pick up on the changes automatically upon launch. What do you guys think of this strategy?

posted on Wednesday, July 18, 2007 at 12:38 PM by Steven

I agree that the Darwinian approach is preferable, and to a certain degree safer. But, I do think the Big Bang approach is useful when you're creating a competing product. Say you have figured out what a competitor is doing, and why they aren't succeeding on a mass scale, building your app in stealth-mode to exploit that problem may bring a more successful launch than if you told everyone about the product pre-launch and allowed your competitor to catch up.

posted on Friday, July 20, 2007 at 8:10 AM by Dan

Isn't this approach applicable for web startups? Or to good extent to Software startups. I understand this blog is dedicated to Software startups only. But things like "getting feedback" and "updating product" are not easy in other domain (eg. semiconductor). And we have to work in stealth mode only. And then use "Big Bang" theory.

posted on Friday, July 20, 2007 at 11:19 AM by Aditya

Dan makes a similar point.. basically, we are facing this issue. but, u gotta hide the tech details.. till the point u really have a strong indication from a potential buyer that he will actually buy your product. Even after such a strong indication is sent by the potential buyer, there is never a guarantee.. u might end up revealing your algorithm/tech details and they will put their hands up and then go ahead and develop it themselves.... and ofcourse, they dont even have to sell it to anyone. If we are talking of selling to competitors itselves, then again the same problems exist. The moment the competitors get the core technical details, thats it... They will build it inhouse and sell it a lot easily because they are already in the same business and most probably have an established reputation... something which start-ups do not have.... Any comments? suggestions?

posted on Friday, July 20, 2007 at 2:47 PM by

I remember the crazy days of the late 1990s. There were VC funded startups paying monthly retainers of $5000 to $15,000 to PR firms. naturally, they all went under a few months later.

posted on Tuesday, July 24, 2007 at 2:56 AM by Smart Startup

Agreed. And when you say "iterate, iterate, iterate" that doesn't just mean the feature set. You may need to iterate your pricing, positioning, value proposition, and overall business model.

This strategy is explained and supported in more detail in Geoffrey Moore's "Crossing the Chasm".

posted on Wednesday, July 25, 2007 at 4:27 PM by Dylan Peterson

I agree with most of the above. I'll take real user feedback over 10 people who were bribed with $75 and pizza anyday. Traditional market research works great when it takes a lot of effort to build/launch ($$$ ad campaigns, rejiggering manufacturing lines, etc.). In the Web world, your best research is getting something out the door. Here are some more of my thoughts on launching products:

posted on Friday, July 27, 2007 at 6:08 PM by Rocky

I agree with what you guys are saying. However, once you pre-market to customers, get beta users, and build, test and launch product 1.0, you will need a solid marketing plan in order to gain mass appeal of the product. The build it and they will come methodology simply doesn't work today in a world of cluttered in-boxes and hundreds of new software products. The software needs to be marketed to a core user demographic who has the largest pain-point met by the software product. If this is done correctly, you will have an avid customer base who will promote your product to other customers who may see ancillary applications for the product. You will be surprised how certain customers see different value in your product, often an unintended consequence of what you designed it for...

posted on Sunday, August 05, 2007 at 1:03 AM by jonathan treiber

Just came across this post and needed to respond. Hope the thread isn't stale yet.

To answer one of Dharmesh's questions in his original post: I am a high-tech marketing professional of many years' experience -- and no, I am not offended or threatened by the organic-growth strategy. In fact, I have spent a fair amount of time during my career fighting the traditional big-company view of marketing.

Why am I not threatened? Organic growth with huge reliance on beta user feedback is not the antithesis of a marketing plan -- it IS a marketing plan! We're not talking here about whether one should have a marketing plan, we're just saying that for certain kinds of products, and in the post-dot-com age, the traditional Madison Avenue big-bucks marketing plan won't work.

In the thread above, Raghu pointed out (and Dharmesh confirmed in his response) that the organic approach is not necessarily EASIER. It is not immediately obvious how to engage potential beta users in an organic strategy. That's what you need a good marketing VP (or consultant) for... to figure this out. But it need not be an expensive process. And you sure don't want a Madison Avenue type in that role!

posted on Monday, August 06, 2007 at 9:19 AM by Carl Strathmeyer

Great post! I like to relate this idea with agile software development. Iterate in small increments and course correct as early as possible. The best (and most successful) software projects I've worked on have this agile mindset both in the development of the product and the business. You can't solve all of the problems at once, and ultimately your customer should be prioritizing the resolution of pain points for you. Hopefully, if you execute the right way, your customers will remain loyal as a vital part of the product's development. Customers that are heard and responded to = happy customers.

posted on Saturday, August 11, 2007 at 2:56 PM by Dan Pickett

Hi there. Agree 100%. I would also think that the first strategy of investing a lot of money now is becoming less and less popular not only among millions of bloggers and start-upers, but also among the companies that have the investment potential. I started by company last year and I had a lot of money to invest, but I chose to invest in people, training and not in product launch. We are outsourcing all of the development, therefore we need to have a strong marketing team. Then they are brainstorming how to make it more efficient that just purchasing a lot of marketing :)

posted on Sunday, September 09, 2007 at 5:18 AM by Vytas Paukstys

This ia great POST! I agree with the latter approach. However, the term Marketing (by definition) includes Sales activities.
What are some thoughts as to launching a Sales campaign. Many blogs discuss the pain of owner selling and then attempting to attract Sales People to start prospecting and driving leads.
Many times they ask about the success of comission only sales people or outsourced services to accomplish this.
I would like to get thoughts about how the marketing plan addresses The First Sales Hire - Out Source or Employee Hire.

posted on Monday, January 14, 2008 at 6:43 AM by Rick McElroy

OK, I’ll join the rest of the stragglers with my comments. I’ve used a Darwinian Evolution approach to launch two startups that eventually filed for IPOs. It is definitely the right approach to get a startup out of the gate, regardless if you are VC funded or bootstrapping. However, eventually you need to accelerate beyond organic growth and start to invest in high growth drivers. If you’ve done your job right during the iterative stage, you’ll have a very efficient acquisition funnel, strong value proposition and a product that delivers on the promise of your value proposition. This makes it much easier to fund profitable marketing campaigns. With the right tracking system in place, you will be able to identify strong ROI campaigns and cut the losers. The only expensive campaigns are those with a negative ROI. My marketing budget at startups has grown to millions of dollars per month, all of which generates a tracked positive return (except for a small testing budget that is generally below 20% of the total). At this point PR is often a great investment, though admittedly less trackable. In a past company trend analysis and surveys indicated that the $20,000+/month spent on PR generated by far our best ROI. I’ve recently joined another early stage VC funded startup and will conserve as much funds as possible as I kick off the evolutionary launch process. Still, in order for it to reach its full potential, we will eventually need to shift into a much more aggressive spending mode.

posted on Friday, February 08, 2008 at 9:56 PM by Sean Ellis (

Not sure I totally agree. I've never done the VC thing, but I've been a part of businesses that did just fine organically for a VERY VERY long time. I think that most of the time people get inpatient or want higher returns. There is the competitive threat. But this does not mean that you can't plod along for a very long time.

posted on Wednesday, November 26, 2008 at 5:40 PM by Adam

I have a question about the advice you gave to Raghu: 
Leave comments on related blogs that contribute to the conversation (and have a link back to your content that is relevant) 
Isn't this frowned upon or considered spam?

posted on Wednesday, April 01, 2009 at 8:36 PM by Kapil

Comments have been closed for this article.