The framing matters more than the wording. Any approach that positions you as the beneficiary requires the other party to do work for somebody they do not know. Any approach that positions you as offering something reverses that, and the conversation starts from a reasonable place.
Establish first that there is an actual overlap. The businesses worth approaching serve the same customers you want, at a different point or in a different capacity. If you cannot state what that overlap is in one sentence, the approach will be vague and it will read as speculative networking.
Make the first ask small enough that declining costs nothing. Asking whether it would be worth a conversation is easier to answer than proposing a meeting, and considerably easier than proposing an arrangement. People say yes to small things and defer large ones indefinitely.
Do the research before writing. Their website, what they actually do, roughly how long they have been operating, and anything specific you can reference honestly. Ten minutes is enough and the difference between a researched message and a generic one is visible in the first sentence.
Approach the person rather than the business where you can. A message to a named individual with a reason for choosing them outperforms one to a general inbox by a wide margin, and finding the right name is usually straightforward.
Be honest about the stage you are at. A first year business approaching an established one gains nothing by implying otherwise, and being clear about it frequently helps. Plenty of established businesses are willing to help somebody starting out, and almost none appreciate being misled about it.
Expect most approaches to produce nothing, and treat that as normal rather than as evidence the method fails. A response rate of one in five on a well researched approach is good. The three or four relationships that come from twenty attempts are what the exercise is for.
Follow up once and then stop. A single follow up adding something useful is reasonable. A second reads as pressure from somebody with no relationship, and it costs you the possibility of approaching them again later under better circumstances.
Then meet in person if the geography allows, because these relationships form considerably faster face to face than over email. The message exists to get to a conversation, and the conversation is where the partnership is actually decided.
Send it from your own address to a named person rather than to a general inbox, because a message addressed to nobody is answered by nobody. Finding the right name usually takes two minutes.
Say plainly what you are asking for at the end. A message that circles the point leaves the recipient unsure what response is wanted, and the default response to uncertainty is none.
Follow up once with something useful and then stop, because a second follow up from somebody with no relationship reads as pressure and closes the door on approaching them later.
Expect four in five approaches to produce nothing, which is normal rather than evidence the method fails. The one that works may produce referrals for years.
Be clear about what you do in the first two sentences, since a recipient who cannot tell what you offer will not read to the end.
Time it sensibly. An approach arriving in somebody's busiest week is declined on availability rather than on merit.