Survey respondent size has relationship with number of questions used.

With all respect, but I don't see how does the appropriate number of domain experts needed for a survey in specific field relates to number of questions in a survey. Any reference or a source for your statement would be useful.

In your framework which you designed, have you covered all the topics of the DW and BI or is it very specific say targeting only "data modeling" or only "predictive analytics"  because your question is kinda generalized, when you say domain experts, I assume you already know whom your framework is focused on , if its a generic framework,  the usual size or rather sample should include 1 or 2 experts, ppl from each sub aspect...

My framework is proposed as a holistic representation of the components involved in Business Intelligence processes. When I say domain experts, I mean technical, business, management and other BI users: BI developers, architects, BI Project and Product Managers, etc.

"When you perform a survey, the intention is to get a representative image about a number of variables or statements within a certain target group or population. Due to practical reasons (too large, too expensive, too time-consuming,…) it is often difficult to interrogate the total population. In that case a sample is used. This is a selection of respondents chosen in such a way that they represent the total population as well as possible."

SAMPLE SIZE

Calculate representative sample size. However, you should check from reliable academic source(!) that this calculator does provide valid size estimations. Basic quantitative survey study methodology book should have such information.

https://www.checkmarket.com/market-research-resources/sample-size-calculator/

SAMPLE VALIDITY

By the way, you have to also make sample analysis by comparing "population sample variables" to "target population variables". For example, if BI domain experts are 80% male and 20% female, you need to later verify that your sample as well as your respondents match to these target population characteristics. If you respondents are 60% of male and 40% of female, your respondents are biased set and do not represent the views of target population. Therefore, you need also to track basic variables such as age distribution, experience distribution, education background distribution etc and compare them to general population. You need to analyse whether your population is biased to "experienced Computer Science experts" or maybe "inexperienced business analysts".

If you are doing fully open survey, you cannot estimate target population well. In these cases, you probably have to do sample validity analyses twice as thoroughly. You need to describe what  kind of respondent sample is, how it might be biased and who they represent. And analyse how your chosen methodology (websurvey in professional linkedin groups etc) has affected to the sample size and characteristics.