B2B prospecting

In B2B, one company sells to another. Several people take part in the decision, cycles are longer and amounts are higher than in consumer sales. Prospecting covers the start of that cycle: spotting the right accounts, finding the right contact, opening a conversation and checking it deserves a meeting.

Two approaches are usually contrasted. In outbound prospecting, you contact people who didn't ask for anything. In inbound, they come to you through your content, your website or a referral. Most teams mix both, because inbound alone depends too much on luck and outbound alone is expensive in time.

The channels are well known: phone, email, LinkedIn, events, referrals. Phone is direct but numbers are hard to get. Email allows volume but follows strict deliverability rules. LinkedIn shows the context before the first contact: role, company, posts, mutual connections. Events and referrals produce the warmest conversations, in small quantities.

Prospecting ends when a qualified meeting is booked. After that, it's sales, carried by a closer or an account executive.

The steps come in the same order whatever the channel: define the ICP, build the account list, find the decision-maker, write the first message, follow up, qualify the replies, book the meeting and track everything in a CRM.

To size the effort you need realistic rates. On LinkedIn, the MimikFlow 2026 Observatory followed 50,930 invitations sent to strangers: out of 100 invitations, 30 are accepted, 25 open a conversation, 10 prospects reply and 1 to 2 conversations end in an observed meeting, or 1.6%. That last number is a floor, because some meetings get arranged elsewhere.

In that cold cohort, an account sending 100 invitations a week, 20 per working day, opens about ten conversations and books one or two meetings a week. Warmer sources change the math: people who had reacted to a post booked a meeting in 18.7% of cases, against 4.6% for strangers found through search.

Measure each step separately: acceptance or pick-up rate, reply rate, meeting rate, show rate, close rate. A single overall number doesn't tell you where things break. If invitations get accepted but nobody replies, it's the message. If nobody accepts, it's often the target.

Then break it down by segment and by source. One campaign can hide an industry that replies twice as well as another.

Give replies time to arrive. In the 2026 Observatory, 15.2% of replies came after the first week. Follow-ups matter too: three follow-ups covered 90.2% of observed replies. Judging a campaign after three days with no follow-up means measuring half the work.

A vague target comes first. "SMBs" lets you neither pick the right accounts nor write a precise message.

Then comes pitching in the first message. Presenting your offer in several paragraphs to someone who asked for nothing is talking to yourself. In the 2026 Observatory, first messages under 200 characters got 50.5% replies, against 19.6% beyond 700 characters.

Another common mistake: no follow-up at all, or a follow-up that repeats the first message word for word. And the opposite, insisting after a refusal. A clear no ends the sequence.

The last trap is volume without follow-through. A thousand messages sent, then replies left unanswered for three days, lose more meetings than a smaller campaign handled well.

MimikFlow handles the LinkedIn side of prospecting. The platform finds the profiles that match your target, checks their relevance, sends invitations at the pace your account allows, writes the first message, follows up as many as eight times and replies with the AI Setter until the meeting is booked in your calendar.

WhatsApp can back up LinkedIn when a prospect's number is already known and authorised. For the detailed LinkedIn method, industry by industry, the LinkedIn prospecting page goes further.