Just finished running ContentBot through its paces. Generated 100 blog intros across a few of my usual SaaS topics. Wanted to see if it was actually useful or just another wrapper for the GPT API.
The good: It’s fast. The tone control is decent, better than most. Got a few genuinely usable hooks for tired topics like “email deliverability” and “sales cadences.” Output is clean, no weird formatting you have to fight with.
The bad: The “creativity” slider is a joke. Crank it up and you get the same cliches, just with more adjectives. Saw the same three intro structures repeated after about 30 generations. For the price, I expected less repetition and more actual variation. It’s a time-saver, but not a mind-replacer. You’ll still need to heavily edit.
CRM is a necessary evil
Your point about the repetition after 30 generations is spot on. I've observed the same pattern ceiling with similar tools. The underlying models have a limited repertoire of narrative templates they default to, especially for common business topics. The "creativity" parameter often just adjusts token temperature, which increases randomness but not structural variety.
Have you tried logging the output to categorize the intro structures you saw? I did a similar test last month and found that nearly 70% fell into three patterns: the "question" hook, the "statistic" opener, and the "imagine this scenario" lead. Tracking the frequency made it clear the tool wasn't exploring the solution space, it was just decorating a small set of skeletons.
For a true test of variation, you'd need to run a controlled experiment changing only one variable, like the seed phrase, while holding the topic constant. My hypothesis is that you'd still see the template repetition dominate, which means the value is in that first batch before the patterns emerge.
Data > opinions