Entrepreneurship
They came. Then left.
Getting someone to visit a website is one thing. Giving them a reason to return is another.
A number goes up on a dashboard.
Visitors.
Article views.
Maybe even a satisfying spike after a story gets shared at the right time.
It feels good. When you're building a news site, seeing people use it makes the project feel real beyond your own laptop.
Then comes a less enjoyable question:
How many of them will return tomorrow?
It's something I keep thinking about while building KUMBA.
The first click isn't the whole story
A person can arrive through Google, social media or a link in a group.
They read the article, find what they need and leave.
That doesn't necessarily mean the site disappointed them.
Maybe they simply got their answer.
But if you want to build a publication people read regularly, the first click isn't enough. You need to understand what makes them return.
Traffic says something about visibility.
Loyalty tells another story.
Three numbers worth separating
Impressions mean content or a result was shown somewhere. Not that somebody opened it.
Clicks and visits mean someone came. Not necessarily that they read carefully.
Returning readers can be closer to the relationship a publication builds with its audience, although measurement has limits and privacy matters.
Imagine two stories.
One attracts a large audience through a trending topic. Another gets fewer visits, but readers share it, explore related pages and return the following week.
Which matters more?
It depends on the goal.
That's why a clear goal should come before celebrating a graph.
What I'm trying to build with KUMBA
I don't want KUMBA to be only a machine that adds headlines to a feed.
I'd like information to connect: news, companies, opportunities and useful reference material.
The hypothesis is simple: information becomes more valuable when it helps people discover the context around it.
A reader finds a story about an investment.
They might then want to explore the industry, companies involved or other developments on the topic.
At that point, the publication could offer something beyond one isolated visit.
It's not yet proof the approach will work.
It's something I want to test.
Avoiding bad incentives
When you're chasing traffic, misleading headlines, mass-produced pages and unnecessary clicks can become tempting.
They may inflate numbers.
But trust is harder to rebuild than a dashboard.
For an automated publication, the responsibility is even greater: reduce duplicates, verify what can be verified, make sources clear and avoid generating text only to fill pages.
I wrote about that in Another project?.
Automation should serve the reader.
A relationship, not just a graph
I'm trying to treat every metric as a question.
Which topic solved a problem?
Which article deserved more explanation?
Which page is difficult to read on a phone?
Where do visitors seem to abandon the journey?
Answering these questions takes observation and sometimes conversations with actual readers.
It also requires respecting privacy instead of collecting data simply because a tool makes it possible.
What I'd like to say someday
I could target an impressive number of visitors and call it success.
But I'd rather say people return because they find something reliable, understandable and useful.
A publication isn't built only when people arrive.
It's built when they decide it deserves a place in their habits.
And for me, that second part is far more interesting than a graph going up.
