Both parties agree on the terms sheet through negotiation. That sounds obvious, but the point is there is not one set process. There are many strategies to negotiation and some make more sense than others in the context of a particular deal. It is not uncommon for parties to walk away and come back to the table at this stage, and in fact, at all stages. The "price" could easily move up or down significantly depending on a number of factors. It really comes down to what the seller is willing to sell for and what the buyer is willing to buy for. If there is an overlap, there could be a deal. If not, then no deal. The rest is jockeying for the overlap. Walking away until you get x is one tactic to get to the top of the overlap from your perspective.
Finally, investment bankers and their equivalents in these situations are not necessary, but they can get you a higher price. Their basic usefulness is connecting you to potential buyers, i.e. through their roladex. You, alone, may not know or have access to these potential buyers. By engaging multiple buyers you can get an auction scenario, which is often ideal because usually buyers willing to pay more for something if they are taking it away from someone else in addition to getting it for themselves. However, certain tactics can turn off and ultimately dislodge particular buyers from the process, which could be net negative for the shareholders depending on which buyer(s) walked away. Also, brokers usually take a significant % fee, and if they do not get an auction going, this could just be wasted money.
Be one of the negotiators. Sounds stupid, but if you aren't at the table, you really aren't being represented, no matter how wonderful the high priced help is.
If only that were true! A great many people have psychological difficulty in following through on what they believe to be in their best interest.
Other people can often be better advocates for you than you are. They may demand more for you than you would be comfortable asking for yourself.
Instead, you could try a larger commission of any sale price above X, where X is your low estimate of what your company is worth.
Edit: On the other hand, this tilts their incentive towards pursuing low-probability, high-payout situations. If someone is working with ten different companies, and they all compensate with this structure, getting one good sale and ignoring the other nine might be the most rewarding option. On the other hand, if there are nine companies with the "percentage of sale price" structure and one with "higher percentage of price-minus-threshold," compensation, the latter will get all the attention. So this boils down to a brutal, Darwinian struggle to pay your sales guy the right way. At some point, structuring compensation might waste more time and effort than just selling the company yourself.