Hire B players if they are 'good enough' to do the work, it won't be seen by major clients and it's not 'critical' to the success or failure of the venture. But ...if that's the case why are you even doing the work?
So...be careful...many functions in a start-up are mission critical (i.e. customer service, sales, design, marketing, etc). Err on over-hiring to start. The compounding of many, critical choices on multiple fronts will make the difference.
Never hire b-players for A-jobs. Having said that...most start-ups don't need custom gold-plated, hand designed bathroom knobs. Not all jobs are A jobs. Hire people to slightly exceed the skill requirement and critical importance of the work. Get the best people you can find at each level...within the time and budget you have...but you may only need/want B players for C level jobs.
For example, a premium, fashion brand may need an A logo. A discount based web retailer selling primarily over cell phones...may only need a C logo. And most start-ups are really cash strapped, so equity is part of the deal in getting most A-talent when you are pre-revenue.
So...clearly, there are dividing lines...and we have to get what is 'essential' to our core business and what is 'experimentation, luxury and/or nice to have':
Some questions I might ask myself:
1. Is this person VITAL to increasing the success of the business? Either to boosting the near-term size and/or growth exponentially or decreasing likelihood of failure dramatically. If so...consider equity. But also ask...is this person's skill set non-correlated with the existing team. If not...be careful.
2. Is this project / function core to testing the basic 'engine' and revenue driver of the business? Then...in general, we probably have to get this project to (at least) 80% right (in a business to consumer model)...to be able to begin A/B and focus group testing. If so...go A. Either as contractor or in-house.
3. Is this work going to be seen by a major account and/or am I selling B2B..where anything sloppy could set us back years...or kill us entirely...or can we, instead, control who will see this (with very limited tests on 'less important' clients...or small subset initially)...so even if it's a mess, we haven't done that much damage? So...never hire b players on presentation materials to major accounts. Better not to show anything than to show b work here.
4. How long and expensive will the 'b' player be? In general - the less money and time you have, the more important A players. If you are not sure what you are doing...it's fine to make them contractors. In fact, it's probably wise. But...I'd always try to get 1-2 core partners on board...and make them as A +++ as possible...early on.
5. Is this a 'fixed' need or a 'variable' need. Do I have a solid sense that this area will grow. If you are getting revenues from your core business...but want to test a pivot...and your core guys are committed...do this through a contractor. If you have proven the core revenue model and don't have the man-hours covered to fill your short-term needs, hire an A player - either as a contractor or employee. Contractor would be preferred to start. On every hire...apart from a core 'partner.'