Twitter Engagement Part 1: Moving Beyond Bulk Following

This guest post comes from Adam Green, a Twitter API consultant and author. Adam blogs about Twitter programming at and tweets from @140dev. His latest book, Twitter API Engagement Programming, is available on Amazon in paperback and Kindle editions.

Successful Twitter engagement is generally measured with the simple goal of gaining a high follower count, but true engagement doesn’t end with a follow-back—that is just the beginning. What you really need for success on Twitter is an ongoing conversation with like-minded individuals, folks who will provide informed feedback on your tweets, introduce you to their friends on Twitter who share your opinions and help spread your messages. This series of articles on engagement programming will show you how to use Twitter API 1.1 to move from simply following to truly engaging on Twitter.

It was the obsessive drive for more and more followers that led Twitter to ban automated following tools in its latest version of the API Terms of Service. These tools used a simple model of high-speed churning of followers. They would follow as many accounts as possible for you, then unfollow anyone who didn’t follow back within a short period of time. The result was massive follow spam. Like any form of spam, it degraded the entire follow process, caused people to ignore follow requests, and broke the follow and follow back cycle. Engagement felt hollow.

The ban on automated following shouldn’t be viewed as the end of development for Twitter engagement programming. There are still countless opportunities for developers to enhance the productivity of individual Twitter users and professional social media managers. The secret is a more sophisticated development approach I call Collect, Identify and Engage.

I’ve been building Twitter engagement systems for clients for years, and the model I’ve adopted follows these basic steps:

1. Collect tweets and user profiles for people tweeting with keywords related to the client’s area of interest using the Twitter API. This data is stored in a database with a schema that makes it easy to perform queries identifying the most active and influential accounts for this topic. This collection system is kept running indefinitely, so a steady stream of leads for engagement emerges.

2. Identify the best candidates for engagement by data mining your tweet and user database. For example, you can find any user with selected keywords in their bio, who also tweets with these keywords at least once a week. The Twitter API rate limits can make it difficult to run this type of query in real time, but once the information is all in a local database, the rate limits have no bearing on how much digging you can do.

3. Engage with targets discovered through data mining more efficiently and track all interactions. I could have stopped at step two by following the best leads in the database, but following should not be mistaken for a relationship. A real relationship on Twitter is an ongoing dialog. You tweet something, someone else replies or retweets, you respond, etc. offers very basic tools for this process, and there is no way to review all tweet exchanges you have with a specific account. Direct messaging is even worse, with no way to even search past DMs. These are just a few of the opportunities for developers to add value to their client’s engagement work.

The rest of this series of articles will dig deeper into each of these techniques, and provide links to complete PHP Source Code so you can implement them yourself. If you are impatient to get started, you can find all of the source code for engagement programming at

Be sure to read the next Tools article: Appy Pie Adds Image Sharing Leaders to API Portfolio