Thursday, August 21, 2008

How to monetize a Tweet (update)

This post is an update to my previous posts regarding Twitter and Monetizing Twitter. I have been debating with Don Draper via the Marketing Pilgrim Blog (original blog post by Andy Beal) why I believe Twitter can successfully monetize, and validate their business model.

Don and I disagree, but to be fair we are looking at different metrics, and time frames. Don keeps referring to Twitter's having 3.8 million users. I am looking at Twitter's user base 2 years out, where I can envision them having 10's of millions of users (my guess is 75Million to be exact). Why such a high number? If we look at their current adoption rate unique traffic has gone from around 500,000 users in December of 2007, to an almost 3,000,000 today (August 08). So let's put that in perspective (conservatively) they have grown more then 500% in only 8 months!! You need to also appreciate though that sites like Compete are only counting unique visitors to Twitter. I believe that very few users ever even get to More often then not, Tweets are via IM, Mobile phones, and MANY third party's using the Twitter API. A quick glance at the Twitter Wiki shows no less then over 100 services (software) that have already taken advantage of the Twitter API. New services are popping up everyday, such as Phweet ...Phweet uses the Twitter API to allow you to have conversations, and conference calls with your Twitter friends. My point here is twofold;

1) To estimate that Twitter currently has 3.8M users is likely an underestimate.

2) Twitter has become an Ecosystem that as a result has 100's of other software vendors bringing them millions of additional users, and unbelievable HYPERGROWTH.

Don in his posts refers to stuff like "1% click through rates". I believe this is short sighted, and leads one to believe that Twitter will sell worthless run of site advertising attached to tweets. That is not the advertising value that I believe Twitter will bring to the table. Let's look at a few hypothetical advertising revenue models for Twitter. Keep in mind my assumption is based on Twitter's FUTURE user base, and my assumptions of what that number will be. This following set of numbers is believed to be factual, based on a source close to the company;

Based on Total Users: 1+ million
Total Active Users: 200,000 per week
Total Twitter Messages: 3 million/day

So when I multiply out those 3 million/day Tweets x 75 Million users, I get 225 Million Tweets per day or 82 Billion contextual blurbs in a database. Granted, these are only numbers but at some point the numbers become large enough to be significant. Suddenly Twitter has a MASSIVE database to send out highly target ted ads to. As a user, I would prefer ads that are highly target ted and not just random.

Revenue Source 1)

Let's pretend for a moment I manage the budget for Coca-Cola. What would be the value to me if I could datamine all those 82 billion Tweets for mentions of keywords I believe to be important to me ... As a few examples "thirsty,drink,diet", or competitors names such as "Sprite,7UP,"etc. What if I could then send out a Sponsor Tweet, perhaps the Tweet would be a redeemable coupon. Let us also assume that my keyword search terms turned up over 1 million Tweeters of these terms (again, I realize these are made up numbers but it is just to make a point).

I might be a concert promoter like LiveNation and for an upcoming concert by The Who, perhaps I want to reach out to people who Tweeted about "The Who,Rolling Stones,etc", but who also reside or our in a certain town (remember Twitter will become VERY location aware as more and more smartphones are shipped with GPS)... Wouldn't that be more valuable to me as an advertiser then a blind $1 CPM run of site buy? My point is this;

Large user base + being able to associate them with products/demographics/location/context = More then just a large random user base. This is true not only for Twitter, but any company that has a database of information about their users. The more info, the more you can target, the more you can target, the more you can charge...

In my exchange with Don, I mention that I believe there are 100,000's of verticals going on within Twitterland. For example, today my fellow Tweeterers consist of only people within the Social Media space (that is my vertical today). Last week, I went to the Police concert and at that given moment the Police became my vertical. As a result, Twitter for me is like having a social network I can join on the fly. If my interest one day is sailing, I need only do a search to locate everyone else who has Twittered on that subject. A few "follow" clicks latter, and I have built my own little Sailing social network on the fly. I don't know about you, but just from advertising alone it seems to me there is substantially more here then $3Million per year (from Don's estimate) in revenue.

Revenue Source 2)

A Subscription service sold to corporations that want to monitor the *buzz (good and bad) about their brand. Lets put a realistic price on this service of only a few hundred dollars per year. A VERY doable number. I do not have an actual total number of "brands", but if you use the metric of trademarks, there are well over 500,000 ACTIVE trademarks amongst over 150,000 registrars. Think that is not a 7-8 figure revenue opportunity?

Revenue Source 3)

Freemium... Offer different Twitter options to different users. Perhaps (as MANY other software companies do) give the basic version of Twitter free. Want to make your Twitter location aware? $29.00... Want to have more then 2000 followers (power users & corporations)? $69 ... (I can think of 100's of ala carte options here that users would likely pay for)... Yet another 7-8 figure revenue opportunity.

Revenue Source 4)

Because third party services let me easily form different groups to Tweet different things to...
"I've fallen and I can't get up!" ...partnership with lifealert. $$

Notification Tweets for Alarm companies...

Sell "911" and alert services to Government, Schools & Corporations...

Revenue Source 5)

Do revenue shares with your API partners (a royalty off monies they make from using the Twitter API).

Revenue Source 6)

Port the massive database of Tweet knowledge on top of Yahoo's "BOSS" (open source search), and enjoy a better more knowledgeable search engine and make money via paid search

Revenue Source 7)

Datamine the Twitterbase and provide market research reports to corporate America...again, smells like yet another large revenue opportunity.

Revenue Source 8)

If you look in the bottom right corner of your Tweets, you will notice a hyperlinked comment ie- "from Twitterfeed" ... These tell you what application was used to send the Tweet. Twitter could charge it's API partners to make this comment more prominent (ie - publish it in red). Twitter could also use the Google Adsense method of letting it's partners "bid" for screen real estate to promote their applications better.

Revenue Source 9)

Private label version of Twitter for internal (behind a firewall) secure communication. IBM built their own version of Twitter to use internally

In Summation, when it is about advertising, do not think of conventional current advertising. Location aware, knowledge base,reach and ability to micro-target make their advertising inventory much, much more valuable then what you think of as today's online ads. Their revenue streams do not need to rely solely on advertising, and the 7 or so I just rambled off the top of my head should exclaimate that fact.

1 comment:

jbum said...

>> Based on Total Users: 1+ million
>> Total Active Users: 200,000 per week
>> Total Twitter Messages: 3 million/day

If you look at the message id numbers being assigned in the search results for (which are assigned sequentially), it seems pretty clear that the total twitters per day is closer to 1.25 million per day.

For example, pick a word that is said fairly frequently, such as "palin". Then do this:

Look at the "view tweet" URL on the most recent message:

On mine, it has the number 921,470,217

Now do an advanced search for the same keyword, 24 hours ago. Compare the two numbers. The difference is a good approximation of the number of tweets per day.

What I'm seeing is about 1.25 million on weekdays, dropping to .9 million on weekends.

I'm graphing traffic for some commonly used twitter words here: