Twitter Building Product for its Customers – Brands

I am sorry to break this to you. Despite the fact that you have 15K tweets, 100K followers and call yourself a power user, you are not a customer for twitter. At best the term user describes you and I. At worst we are the raw materials. The real customers are those who pay for its value. So brands that pay for the data, get in front of us with messages and get us to do something are the real customers for twitter.

There is nothing wrong with that. Just making it clear.

Three recent major product announcements from twitter reflect their realization and focus on its customers.

  1. Direct Message Changes –  They removed the 140 character limit on DMs and added group messaging capability. I do not have data on how many DMs are exchanged between non-brand users as  way of communication. Anecdotally I can say all DM I received are promotional messages sent using a third party broadcast tool. I do have data on length of email messages that serve as proxy for DM length.
    We most likely will use twitter DM only as a mean to communicate privately with those with whom we do not have other connections – i.e., no email address, whatsapp, phone for text etc.
    There is no reason to believe DM pattern will be any different from emails. A recent analysis by The Journal says our median length of emails sent from mobile phones is 20 words. At an average word length of 6 (with spaces), 140 characters is more than enough. And more and more of our usage is moving to mobile and tablets from laptops.

    So the the only possible explanation for a DM feature couched as user experience enhancement is to help brands send longer messages.

  2. Moments – It is positioned as a way to get those new to twitter get to know it. For “power users” a way to find hand curated happenings. Do we really lack ways to find what is happening and popular? The different News apps use algorithmic ways to serve us what is popular and (remotely) relevant. That is a crowded field and some of popular darlings like Flipboard are being crowded out by platform players like Apple. So the only possible use case is for brands to insert “Paid Moments” and have Moments integrated into the timeline.
  3. Poll – This is a annoyance to all of us, users. And if you are a marketing research practioner you will be up in arms about how silly it is. It is as scientific as Fox News poll asking its viewers if Hillary Clinton is hiding something. Don’t even bother publishing your results from a twitter poll you did on your followers. The only valid use case is for brands to get people to interact with it. The results the brand will collect are immaterial but a crazy enough poll question will get the question and the brand in front of more people. A simple click of button makes it a low calorie effect than typing a reply. Brands will use it simply as awareness generation tool. Our timelines will be filled with such silly polls.


Overall I see three new product features that are not about users but about brands. That is fine as twitter has to build products for its customers. The focus is on customers not users. That explains the declining user numbers.


Who Do You Create Value For And Why?

This is a guest post by my friend and classmate Rachel Wolan. I worked alongside Rachel in helping run the Berkeley Digital Media Conference, >play. Rachel is the founder & CEO of YadaZing – product maven, ex-Facebook and ex-SAY Media. She shares her thoughts in her blog and tweets at @rachelwolan.

Have you written your guest post yet?

I am the founder of, a media platform for English language instructors (producers) and students (consumers). We maniacally focus on the question, “who do we create value for and why?”

In every platform business, you need to solve the Chicken/Egg problem AND make sure that producers produce. On our platform, we are focusing on the underserved but small segment, English instructors. We give them tools to engage and interact with English students, the overserved but enormous segment.

We are making a bet that our early adopters will be part of the 1% (of the 90/9/1 rule) who will create great content; this is our gateway to the 99% who consume the great content.

Who do you create value for and why?

What are Twitter’s jobs-to-be-done?

This is a guest post by Hutch Carpenter. I met Hutch when he was running products for Spigit. He is one of rare product management leaders who starts with customers and their job to be done over myriad product features. Hutch now does Innovation Management Consulting for startups and enterprises. You can read his thoughts on innovation in his blog and follow his tweets at @bhc3. Hutch’s motto is, ” You can’t wait for customer insight. You have to go after with a club.”

Will your thoughts be featured next in this guest post series? Write yours now.


Jobs-to-be-done asks this: why did customers hire your product? “Hire” means they picked the product to satisfy some need or want. To think this way is to empathize with customers, better understanding growth opportunities. Two things to know:

  1. Jobs are relatively unchanging over time
  2. Products to fulfill jobs are not the same as the jobs themselves

Take Twitter: how’d anyone ever think *that* would be successful? Because it fulfills these jobs:

  • Learn new information
  • Share interesting stuff
  • Talk with others
  • Participate in major events
  • Connect with like minds
  • Establish professional reputation

We used to MySpace, email, write letters-to-the-editor, etc. Now we tweet.

Today we use reasons, Or do we?

What is a predictor for a political candidate’s success? It is in the millions of tweets about the contest.

What will get more customers to buy? It is in getting them to like your brand’s facebook page.

How can you get your message across? It is by getting all your employees tweeting in unison.

What is a predictor of getting your tweet retweeted? The secret is in using more adverbs (or is it adjectives)

When Stephen Hawking wrote about our early models about the universe he wrote,

Ignorance of nature’s ways led people in ancient times to postulate many myths in an effort to make sense of their world. According to Viking mythology, eclipses occur when two wolves, Skoll and Hati, catch the sun or moon. At the onset of an eclipse people would make lots of noise, hoping to scare the wolves away.

If you look at it Bigdata of these days is no different from what those Vikings in early days did trying to understand eclipse. Bigdata or  not is defined not by volume (ok velocity, variety) but by our ability to process with tools at our disposal and how advanced our mind and reasoning is.  It is not farfetched to say,  our early ancestors when bombarded with cosmic data saw that as Bigdata problem, trying to make some sense out of it with myths.

Hawking was very optimistic and forgiving by adding that

Today we use reason, mathematics and experimental test—in other words, modern science.

Today we have no scarcity of data, we have abundance of it – billions of tweets, facebook updates, instagram pictures – all volume, variety and velocity. Do we really use reason,  (right )mathematics and experimental test to make sense of this social media noise?

May be Hawking is true in case of science but not in social media science, data science or in so called science of marketing. If not would be seeing any of those statements we saw in the beginning of this article, propagated as “scientifically proven” methods and retweeted by millions?



What is worse than a meaningless Social Media metric?

The answer is, meaningless social media metric presented in the form of a infographic. And it comes to us via MediaBistro from, not Social Media Gurus, no not witch doctors, and if you guessed Social Media Scientist that would be a good guess  but not correct, because it comes from SocialBakers.

If you have not figured out the flaw in the metric (shown left) or you are one of those 311 people who tweeted the link, I will try to point out the problems with bakers’ recipe.

The bakers give us a simplistic formula for what they arbitrarily define as Average Engagement Rate. This is not a model derived based on theory, data collection and experimentation but simply a formula thrown together with a mix of arithmetic operations. Disregard for a minute how it is computed (which is beyond ridiculous) and ask the following questions:

  1. What does this really mean in the context of your marketing spend?
  2. Why is this important?
  3. What is the marginal benefit from a change in this metric?
  4. More broadly, what does the curve look like ( Revenue = f(Average Engagement Rate))
  5. What is the cost of moving the AER, if at all one could?
  6. Are the AERs of Facebook, twitter etc additive?

Well keep asking as you are not going to find any such answers from the eye candy infographic.

Now to the computation.

First the units of this metric. It is stated as a percentage. And percentages are? Dimension less. The numerator at best is a dimensionless ratio and at worse has a complicated unit of Interaction/Activity. The denominator has units of number of people – fans, followers etc.  In other words the ratio is not a dimensionless quantity because you are dividing  a quantity that is NOT number of people by a quantity that is number of people. So how can anyone simply multiply this number by 100 to state this as a percentage? Like Baker’s Dozen, they should call this Baker’s percentage.

Second the ratio is specifically designed to show as low a value as possible and hence the possibility for improvement and potential sales. The numerator is divided by total number of followers (or fans) a brand has. So larger the number of followers, smaller the magical AER. The thermometer in the infographic tells us the maximum any brand currently has is 0.7%. This also explains why they chose to multiply by 100 and call it a percentage. Otherwise, 0.007 would look too awful for anyone to pay attention.

Third, this is indeed genius move in targeting, after all those brands with millions of followers likely have higher marketing spend and hence are likely to divert some of it to improve their measly AER.

There you have it, yet another pointless and wrong Social Media engagement metric, presented as a stunning infographic that not surprisingly found many takers.

Final note, if you are tempted by any of the social media engagement metrics that talk about anything except dollar values you can cure that temptation by reading Ron Shevlin‘s book Snarketing 2.0.

My First GitHub Commit – Select Random Tweets

I have seen several reports that collect and analyze millions of tweets and tease out a finding from it. The problem with these reports is they do not start with any hypothesis and find the very hypothesis they are claiming to be true by looking at large volumes of data. Just because we have  Big Data, it does not mean we can suspend application of mind.

In the case of twitter analysis, only those with API skills had access to data. And they applied well their API skills to collect every possible tweet to make their prediction that are nothing more than statistical anomalies. Given millions of tweets, anything that looks interesting will catch the eye of a determined programmer seeking to sensationalize his findings.

I believe in random sampling. I believe in reasonable sample sizes, not whale of a sample size. I believe that abundance of data does not obviate the need for theory or mind. I am not alone here. I posit that the any relevant and actionable insight can only come from starting with relevant hypothesis based on prior knowledge and then using random sampling to test the hypothesis. You do not need millions of tweets for hypothesis testing!

To make it easier for those seeking twitter data to test their hypotheses I am making available a real simply script to select tweets at random. You can find the code at GitHub. You can easily change it to do any type of randomization and search queries. For instance you want to select random tweets that mention Justin Bieber, you can do that.

The script has bugs? I likely does. Surely others can pitch in to fix it.

Relevance not Abundance!

Small samples and test for statistical significance than all the data and statistical anomalies.