If we have a presenter who is not confident in skilled in public speaking situations then a script and an autocue / teleprompter is required. And to be honest they help get the job done fast even when the presenter is a skilled public speaker.
When presenters write scripts they can tend to be a little ‘written’ which comes off unnaturally when read to camera. We never underestimate just how different written and spoken forms of English are.
To alleviate this we make sure that all possible spoken contractions are contracted in the text itself, and not left for the presenter to amend on the fly. (ie. it is = it’s, we are = we’re). English of course is not a stress-timed language like, say, Italian and so we can choose to place the stress wherever we like - and we've found that it is not as predictable as you think when you have a presented making those decisions on the fly while reading an auto cue with a bunch of cameras pointing at them - it can help therefore to capitalize key words that carry where the MAIN sentence stress should fall in the context, or in fact where the main STRESSES should fall in the CONTEXT.
These days, teleprompers designed to work with a tablet device can be rented cheaply from an A/V store. We learned to avoid using the prompters that work with iPhones after getting some very squinty results.
If we have a client who is exceedingly well versed in the topic, an expert who may have taught or coached the subject many times before - we use an interview format which allows for a more natural and less strictly precise flow of language than required in scripted contexts. We either have the presenter sit at a table with an interviewer, coffee shop style so the viewer feels like they are sitting right there are the table with them. There is more room for manoeuvre. With no requirement for a script this can save the presenter considerable time in prep. The only thing we pre-script are the questions so that the presenter can be prepared. We want the online learner viewing the video to feel they are part of the conversation so try to create a ‘coffee-shop’ format around a small table -the camera positioned to give the viewer the feeling they are at the table too.