Research

Original data the market has to cite

Two studies a year, each with a published method and dataset. Original research is the most defensible asset a firm can build: it compounds, it cannot be faked, and it earns the links that carry authority to the pages that sell (Section 16.1).

How a study is run

The same five steps each time. If a step cannot be done properly, the study is delayed rather than published thin.

  1. Define the question

    One question a Kenyan business owner or journalist would actually ask, narrow enough to answer with evidence.

  2. Draw the sample frame

    How the businesses or prices were chosen, from where, and what was excluded. Published before collection starts.

  3. Collect

    Automated measurement where it can be automated, collected quotes and anonymised submissions where it cannot.

  4. Analyse

    Plain descriptive statistics. No modelled projections dressed up as findings.

  5. Publish the dataset

    Findings, method and the underlying data together, free to cite with attribution.

Why we publish it

Kenya has very little public data on what websites cost or how they perform. That gap is filled by guesswork, and businesses pay for the guesswork.

  • It can be checked

    Method and sample published alongside the numbers, so anyone can disagree with us on the evidence.

  • It compounds

    A study repeated each year becomes a trend line, and a trend line is worth more than any single snapshot.

  • It cannot be copied

    Anyone can rewrite a guide. Nobody can rewrite a dataset they did not collect.

Where to go next

Until the studies land, the guides and the published prices carry the detail.

Journalists and researchers: email us for early access to a dataset when it publishes, through the contact page.