After the Techcrunch Bump

I see many consumer Internet pitches these days where the basic marketing strategy is to (1) get covered by Techcrunch, (2) get tens of thousands of users from the "Techcrunch Bump", and then (3) "grow virally".  While a positive Techcrunch review has the potential to send thousands of consumers your way, it does not represent a marketing plan.  Munjal Shah at Riya found this out after the launch of Riya back in 2006, when he wrote about "the cocaine like high and subsequent crash of the Techcrunch effect":

The Techcrunch article got put on Digg and read in thousands of feed readers and viola... the Techcrunch effect begins. Michael's blog is the single more effective vehicle to get the word to the online blogosphere about new technology companies on the planet..The unfortunate fact was that the initial hammering [our servers] took was just not the reality we would see later. The number of photos uploaded per hour began to fall and then stabilized near the end of the 22nd at around 25,000 photos per hour and would continue to fall for weeks to come. - Munjal Shah

So while a positive reception from the blogging community is valuable -- and can generate a lot of initial activity/interest and a nice looking Alexa chart -- it is not the only ingredient in your ultimate marketing success.  When I see a post-launch consumer Internet startup, I basically look for a few simple things:

1)  Usage Growth -- how many unique users are visiting/engaging with your site and product, and how is the rate of growth evolving over a several week period of time.  I also look at the source of this growth -- is it scalable, repeatable and systemic?  Is it event-driven (ie, PR)?  Is it organic or driven by marketing (ie, is the company buying growth via Adwords, etc)?

2)  Virality -- So many people misunderstand virality.  Virality is not "word of mouth".  And having a product go viral is not easy -- nor is it something you can just "sprinkle on a product" after creating it.  If making a product viral was as easy as adding a "share with your friends" button, there would be no reason for the $100 Billion advertising industry.  (I can see companies asking themselves -- "let's see, should we spend millions on advertising...or should we just add virality...Hmmm").  I believe a viral product is one where a consumer's basic usage of a site/product brings new users (and therefore additional utility) to the site/product.  Facebook, LinkedIn and Paypal are all great examples of viral products.  If you're pitching your business, you should know your viral coefficient.  That is, how many new users get added virally from each additional user.  And if you can get your viral coefficient greater than 1.0, then you've built something really special. 

3)  Engagement Level -- Do your visitors actively engage in your site?  How long are they there for?  How many pages do they view?  What is their user experience like?  One of the easiest ways I've found to evaluate a company's engagement level is to have them (temporarily) share access to their Google Analytics account -- this gives us the ability to get the data/insight we'd need without having to bother them to run each and every report. 

4)  Repeat Usage -- User retention tends to be an area where people pay the least amount of attention, but I think is one of the most important to monitor.  Specifically, how often do people come back to your site.  While there are a lot of different ways to measure retention, my preferred way is to look at a cohort analysis.  Say you've had your site running for five months -- you now have five "first month cohorts", four "second month cohorts", three "third month cohorts", two "fourth month cohorts" and one "fifth month cohort".  And you can see, what percent of your users come back in each subsequent month.  A simple chart is below.

You can also plot it out graphically -- I've attached a generic Cohort Analysis Excel document.  From this data you can learn a lot.  Not only do you see how many people are returning this month, but you can see the trends over time.  For example, in this model spreadsheet you can see that while the site is still just retaining a small percentage of their overall users, the rate of retention has gone up by over 250% over the course of the year.  And while this example cohort analysis is shown by month, immediately post-launch I'd recommend that you create and track cohorts by week.

Comments

[Eyal Hertzog]

Great post. Thought I think that there is a bit more to "virality". Google Search was one of the most viral product ever. Even consumer product like iPod and BlackBerry owe much of their growth to their viral effect. I think that there is some level of excitement that one can experience from a product that would simply make him pitch it all around. There is a sort of personal gain by pitching a product that you like to a friend. Obviously, the product needs to have an extraordinary value..

[Tony Wright]

Am I missing something on scribd? How do you actually DOWNLOAD the spreadsheet (or was that not your intention)?

I'd love to have a copy.

[James Siminoff]

Josh,

As you know we had a post on TechCrunch yesterday. We also were posted on TechCrunch last year around the same time. Being both a reader of TechCrunch and a person with a company that relies on viral news growth I think there are no silver bullet posts out there any more. I can tell you that the impact of a post last year was much larger and that was with the blog having fewer readers. This is for 2 reasons I believe. The big blogs post so many things that the impact is less. Also we are in a time where there are so many new things coming out that the consumer is fairly ditracted so it is hard to get them to move from the article or blog to action. I think that it is still an important part of the plan but one which I would not want my money riding on.

Jamie

[Jeffrey McManus]

I'd rather have 1,000 qualified new users who are really motivated to use my site than 10,000 drive-by users referred to us by some blogger.

This notion of cohort analysis seems very useful -- we are going to add it to Approver.com's internal metrics.

[Erick Schonfeld]

Hey, we only get the initial word out. The rest, as you say, is up to the startup.

[Jeremy Toeman]

Good post Josh, thanks to these non-"marketing plans" my firm's getting a lot of work! :)

The other reality here is the "bump" effect from TC and even Digg is much less dramatic today than it was even six months ago. It seems like these sites are good for tech-industry awareness, but Jeffrey's comment above is very appropriate regarding quality users.

[Alx Klive]

Great post.

The concept of viral coefficient is spot on, and a good phrase to describe it. We've been calling it 'naturally organic growth equation' - the idea, if a site has no marketing or ad spend, does it naturally, organically grow over time?

You can work with a site that is even or slightly negative in this regard. If it's a positive number, you're laughing.

[Richard Giles]

That's very true. I never knew why I did a marketing degree until I launched a startup ;).

The comments about TC's referral traffic waning is very true. I've noticed it, as well as a couple of other startup CEOs that I have spoken to.

The impact is close to insignificant these days for startups, probably due to saturation.

[Ahikam Kaufman]

I humbly think there are 2 or 3 level of "virality"

1. Virality level 1 - ie Skype. I can't enjoy and use Skype if my friend is not on it.

2. Virality level 2 - I need to invite friends but they need to set themselves up in the system which is more cumbersome. This would be linkedin, maybe facebook, matching sites etc

3. Virality level 3 - like hotmail used to be - you would see the link to subscribe to their free email but other than that no immediate value (for your friends) in using this

[Jim Keenan]

I think you hit it on the head. What does surprise me however is how many V.C's have NOT asked us for this information. They all seem to be focused on the "1 million uniques". We have met with several "big" V.C's and none have asked us about our user engagement. Our users are very engaged, almost 2 visits a day for an avg. of 8-10 min. a visit. Exclude the drive-byes we get from stumble and it is more like 12-15 min per visit or 24-30 min a user. Because of this we have quite high page views per user as well. I have always felt this was a VERY valuable metric, but until now, know one has seemed to care. Nice post, thanks!

[Matthew]

Josh:

Enjoy your posts; and I am keen on your investments.

In terms of this post, I think this is more than an excellent point that was waiting to be said.

Furthermore, I think something often overlooked with a TechCrunch post is that it will certainly pop traffic/users on your site, but *qualitatively they are probably not the right ones.

They are, more than likely, not looking at the value proposition of the product but merely researching or perusing page flows and funcationality.

*Qualitatively they are not the right users. So building functionality of their patterns is probably not wise. It is best merely to go to market with the value proposition promote it to a general or subjective target audience and then insert viral hooks.

A concurrent point to this: qualitative review of traffic is not representated by Google. Coming from the lead generation and comparison shopping space, overall conversion means little to me. I am interested in conversion of high quality (as I or more appopriately the data dictates).

Thanks for the blog; I enjoy reading it.

[Renee Blodgett]

Great post Josh and it's about time for a reality check.

When I first moved to Silicon Valley less than four years ago, my first two years were spent listening to potential clients who all wanted their product launches at Arrington's house OR a fat hit on Techcrunch around launch time and 'pretty please - every couple of weeks after. Mike even wrote an amusing post about a CEO who came to him nearly in tears if he wouldn't cover them. Some CEO.

Many of these client prospects were targeting vertical markets, like finance, real estate or law or very mainstream consumers who still listen to traditional radio and read syndicated newspaper columns.

While I'm a huge believer in the power of the blogosphere, remain a loyal reader of Techcrunch and other blogs in its category and love the value they provide (because they DO provide a value), getting a hit in a large tech blog isn't going to make or break your company.

What remains key is building a business through bringing true innovation to market and listening to and being loyal to your customers.

[Sean Ellis]

Josh, this is probably the best post I’ve read on marketing a startup. In the two startups I’ve taken from launch to IPO “hope marketing” was not a factor in our success. I completely agree that “scalable, repeatable, systematic” marketing programs are the key to driving sustainable growth. The marketing role is all about discovering/developing these programs. This requires the right tracking systems to give the best possible insight into the ROI of every marketing dollar spent. Google Analytics can get you part way there.

The rest of your post is really about the levers that affect ROI. While most marketers focus on the acquisition cost of a new user, equally important are the lifetime value and viral coefficient. Lifetime value and viral coefficient are directly affected by “engagement level” and “repeat usage.” As the lifetime value and/or viral coefficient increase, the amount you can spend to acquire new users also increases. In fact, if your viral coefficient exceeds 1.0, every dollar spent on customer acquisition is spent profitably since you will collect dividends on that investment in perpetuity (user get user chain never dies). While it is extremely rare that a product will have an overall viral coefficient exceeding 1.0, it is less rare that an individual segment will. When this happens, all marketing dollars should be poured into that segment and you have a big winner on your hands. Still, it is much more likely that you will have to build a customer acquisition engine that requires constant tweaking and investment to deliver sustainable growth.

BTW, I'm currently busy applying this approach to Xobni.

[Dr Vidy Potdar]

Dear Josh
Intresting article.

We here in Australia are working on assessing users contribution in online communities.

We have developed models for user contribution in Forums, Wiki and currently working on blogs.

I thought it may be intrested to the community here.

Dr Vidy Potdar
http://drvidy.wordpress.com
http://www.debii.curtin.edu.au/~vidy