Global MSP CommunityMSP growth · Technology partnerships

Research agenda · Proposed study design

How will buyers discover their next MSP?

A transparent proposal for examining provider visibility, representation and shortlisting across search, AI answers and the wider technology ecosystem.

An aperture reveals selected forms within a broader field of providers, illustrating discoverability without implying measured market coverage.

The question before the measurement

A provider can have a technically accessible website and still be absent from the choices a buyer considers. Discoverability concerns that wider process: becoming visible, being represented accurately and entering a credible shortlist. This proposed programme asks how those stages differ across search, AI assistants, directories, referrals and technology ecosystems.

The immediate objective is to understand mechanisms and variation, not to publish a universal visibility score. A question about a local support provider and a question about a regulated-industry specialist may draw on very different evidence. They should not be pooled without an explanation.

Four questions to investigate

  1. Which provider characteristics appear in buyer questions and in the answers returned by discovery systems?
  2. Which owned and third-party sources are used to describe or recommend providers?
  3. How consistently are a provider’s services, geography and specialisms represented?
  4. How do MSP leaders interpret changes in discovery and adjust their marketing, evidence and partnerships?

These are research questions. They do not establish that AI assistants have replaced other acquisition routes or that any particular website intervention improves revenue.

A proposed design with distinct evidence streams

Public-web observation. Define a bounded set of provider-selection questions, countries, languages and provider types. Record the exact prompt, platform, visible model or interface information, date, location assumptions and complete answer evidence. Store cited URLs and check whether each cited page supports the relevant description.

Repeat observations. Re-run a declared subset to examine variation. A single answer is an observation, not a stable ranking. Record failures and empty responses rather than removing inconvenient cases. Avoid comparing results produced by different prompt wording as if only the website had changed.

Practitioner accounts. Separately consented interviews can explore how owners interpret buyer behaviour and make commercial decisions. Such accounts explain experience and judgement; they do not independently prove a platform’s causal effect on sales.

Comparative cases. Where permission and suitable evidence exist, compare provider positioning, page accessibility, external references and observed discovery. Keep the scope small enough to inspect the evidence behind every interpretation.

Sampling is a substantive decision

The sampling plan should specify provider size or business model, service focus, geography, inclusion criteria and recruitment route. Convenience access through a community or directory creates selection effects. Participating firms may be more interested in discovery, more digitally active or easier to identify than non-participants.

The Global MSP Directory may help identify potential cases, subject to its coverage limitations. It is not a verified census, a random sample or a denominator for estimating the global market. Report how cases were found and why they were selected.

What the analysis should and should not claim

A useful observation record separates a brand mention, a direct recommendation, a citation and an accurate description. Those are different outcomes. It also distinguishes the provider’s own page from a directory, review, media story or partner reference.

Compare patterns across declared cases and repeated observations. Where explanations remain ambiguous, retain the ambiguity. A website revision followed by a new mention cannot isolate the effect of that revision from retrieval changes, source updates or answer variation.

Google’s current guidance retains established search fundamentals for its AI features. OpenAI documents separate controls for search crawling and model-training crawling. These establish platform requirements and distinctions; they do not guarantee that an accessible MSP page will be recommended.

Programme status and participation

This is a proposed study design and research agenda, published to make the approach inspectable. No completed participant sample, original dataset or empirical findings are represented here. A future study should publish its actual protocol, collection window and limitations alongside its results.

Community research and separately governed doctoral work must remain distinct. Participation in one does not automatically enrol a person in another or grant permission for marketing. Recruitment requires the appropriate study information and consent process.

Email your enquiry →

Opens your email application. This does not confirm a booking or subscribe you to marketing.

Scope & limitations

This document describes intended questions and methods. It is not a completed study, an institutionally endorsed protocol or evidence that a particular intervention causes visibility gains.

Sources & further reading

  1. Google Search: Google Search: AI features and your website · Live documentation checked 2026-11-15
  2. OpenAI: OpenAI: Overview of OpenAI Crawlers · Live documentation checked 2026-11-15