Methodology

Purpose

To understand how people and organizations in Ghana and across Africa connect today, what they spend, what goes wrong, and what conditions would need to be met before any new connectivity option — satellite or otherwise — is worth adopting.

Design principles

• Neutral wording; satellite is compared alongside mobile data, fibre/fixed wireless, community Wi-Fi and “keep what I have”.
• Prior awareness is measured before a short, balanced explanation that includes satellite’s limits (equipment, power, sky view, weather, delay, data policies, cost). Usefulness is asked only after.
• Interest is kept separate from ability to pay: price questions are hypothetical, separate one-time equipment from monthly service, and always include “would not consider at any price” and “not sure”.
• Hypothetical price ladder: Ghana respondents are randomly shown one of three monthly/equipment price levels (GHS 150/1,500 · 300/3,500 · 500/6,000). These are research scenarios, not real offers or market prices.
• Skip logic shows organization questions only to organizations, and price bands in GHS only in Ghana; other countries enter amounts in their local currency.
• Every question is versioned. Changed wording creates a new version so results can be compared honestly.

Limitations

• Not a representative sample. Respondents are self-selected from whoever receives the link. There is no sampling frame, so results describe respondents only, not national populations.
• Recruitment-channel bias. An online survey under-represents people who are offline — often those with the greatest connectivity needs. Offline or assisted interviews are recommended to reach them.
• Stated vs. actual behaviour. Willingness-to-pay answers usually overstate real purchasing. Treat them as directional.
• Duplicate protection is limited. To preserve anonymity we only prevent accidental double-submits; we cannot stop a person from deliberately answering more than once.
• Completion rate is not measured because that would require tracking people who start but do not finish.

How to read results

• Every chart shows its base (n) and how many did not answer.
• Groups with fewer than 30 responses are flagged as small samples; differences there are likely to be noise.
• No market-size projections are produced from this data.