I think this means we have to address some potentially specious statistical reasoning here. Assuming that a point cloud of pre transaction questions gives you the information to ascertain the career trajectory between great engineers and the "best" engineers is likely a fool's errand. What's the data source? Is it your technical screen? If there is any possibility that your technical screen doesn't track reality accurately, then your ranking metric is probabilistically very likely to be inaccurate. This is the risk of trying to quantify skill and potential levels before you try to qualify them properly.
Most recruiting platforms work on a transaction processing fee on a per placement basis. Triplebyte is no different. The entire product is set up for a candidate to land a job. It is not set up for a candidate to structure and guide their entire career. The simple way to think of a transaction is as a point: point score on an assessment test, point score on a technical screen, point offer extended or not. You see these things being measured in this blog post, but that doesn't tell the full story to a hiring manager. If you're hiring an engineer that's actually growth oriented, you care about looking at the first and second order derivatives of all of these things, and you'll see them on a project to project basis, or on a person to person basis based on how they interact with their team and other teams.
I want to repeat myself and restate that most recruiting platforms work on a transaction processing fee on a per placement basis. It's erroneous to try and draw the wrong kind of conclusions from data points gathered from enabling this transaction, especially when it's just 1. When a candidate receives multiple offers that they're interested in, it's very likely that they'll make a holistic judgment based on a combination of factors, and their gut. Is the company "good" or "sketchy?" Do they feel chemistry with the people they'll be working with? Are they stimulated by the work the team will be doing over the next year? The truth is, for most talented engineers at early stage companies, any company that makes a desirable product and has substantial growth left in building and scaling its product will allow for growth in literal terms. The limiting factors there will likely be the team, if anything. If there's juice left to be squeezed from building further product edge, and the team is good, the org is good, and the leadership is good, it's very likely that an engineer will receive more than one opportunity that is "good enough" -- good enough for them to exercise the upper limits on how far they can grow.
It's easy to pay lip service to offering growth, but it's hard to actually do. In many cases, companies shouldn't be trying to offer that with a straight face. In growth stage startups, growth as an engineer is bound to the combination of the growth of the company and the growth of the product. In most cases, if you want to hire any kinds of the best employees, you need to offer growth. But, your business model, company growth and stage of company and culture all have to offer that. That's the hard part. It's something that needs to come from the executive level as well as good timing and a ton of foundational work.
If you can offer growth, that's great. It will make it easy to hire the best. But, most companies are not in the position, and the ones that are certainly don't need this advice. So, this isn't really useful. Offering growth is not something you can turn on and off. You need to build a great company and build a great product that people want and which a great business can or is built around. If you can do that, expanding the engineering team becomes a matter of logistics instead of intractables.
I think TripleByte is a great idea, and it's at an interesting point right now. But to use your data to answer this question requires looking at things longitudinally, and that could be hard unless TripleByte becomes more of a career management platform as well as a two sided hiring marketplace. Having used Hired, Angellist and TripleByte before on both sides, I'm definitely really curious about seeing where things evolve.