Pipeline maths is a chain of rates: accounts to contacts, contacts to replies, replies to meetings, meetings to opportunities, opportunities to orders. Multiply through and you get orders; divide back and you get the accounts you need. The hard part is the rates. Published cold-email benchmarks count different things, so use them as a range and your own quote history for the rest.
What is the chain?
Six stages, each turned into the next by a rate.
| Stage | Turned into the next by | Where the rate comes from |
|---|---|---|
| Accounts | Contacts per account | Your choice: who at each firm buys or specifies |
| Contacts | Reply rate | Published benchmarks at first, then your own sends |
| Replies | Meeting rate | Your own replies, once you have enough |
| Meetings | Opportunity rate (meetings that become a quote or tender) | Your quote history |
| Opportunities | Win rate | Your quote history |
| Orders | Average order value | Your order book |
Two of the six rates are about outreach. The other three are about your sales process, and you already have the data for them.
What do reply-rate benchmarks measure?
Two published benchmarks are worth knowing, and they are not measuring the same thing.
| Benchmark | Figure | Divides replies by | Counts |
|---|---|---|---|
| Smartlead | Median 0.74 percent; top quarter 1.43 percent; top 10 percent 2.63 percent; bottom quarter 0.37 percent | Contacts | All replies, from 850 million+ emails, senders with at least 500 emails |
| Instantly | Average 3.43 percent; top quartile 5.5 percent+; top 10 percent 10.7 percent+ | Emails sent | All replies, including to follow-ups; data 1 January to 18 December 2025 |
Source: Smartlead, Average cold email reply rate in 2026, no publication date shown, checked 1 October 2026. Instantly, Cold Email Benchmark Report 2026, checked 1 October 2026.
Smartlead’s median works out at one reply per 135 contacts. Instantly’s figure looks higher partly because the base is different: a four-email sequence to 100 contacts sends up to 400 emails, so the same replies divided by emails rather than contacts give a different percentage. That is why the two are never averaged, and why any reply rate you see quoted needs its denominator stated.
Both also count every reply. “Not interested”, “wrong person” and “take me off your list” are all replies. A reply rate is a measure of attention, not of demand.
How do replies become meetings?
Split replies into positive and the rest. A positive reply agrees to talk or asks for information; everything else is a no, a not now or a redirect. Then count how many positive replies turn into a meeting that actually happens.
We plan on one meeting for every four replies. That is a KS Media planning assumption, not a published benchmark and not a measured result: we have no client outbound results to publish yet. Use it to start the sum, then replace it with your own figure as soon as you have enough replies to count from.
Why are your own numbers the best benchmark?
Because the later stages are where engineering firms differ most, and nobody publishes them for your market. A switchgear retrofit specialist and a UPS maintenance firm may convert meetings into quotes at very different rates, and both would have good reasons.
You can pull the three rates from records you already keep:
- Opportunity rate. Quotes or tender responses raised in the last two years, divided by first meetings with new accounts in the same period. If you don’t log first meetings, count from the diary.
- Win rate. Orders won divided by quotes submitted, in value and in number. Leave out repeat orders from existing customers, which convert differently.
- Average order value. The mean of won orders from new accounts. If a few large orders skew it, use the median as well.
A worked example
This is the pipeline calculator’s default case, with the three published reply rates beside it. Everything except the reply rate is held the same.
| Line | KS Media planning target, 2 percent | Smartlead top quarter, 1.43 percent | Smartlead median, 0.74 percent |
|---|---|---|---|
| Target accounts | 300 | 300 | 300 |
| Contacts (3 per account) | 900 | 900 | 900 |
| Replies | 18 | 12.87 | 6.66 |
| Meetings (1 in 4 replies) | 4.5 | 3.22 | 1.67 |
| Opportunities (30 percent of meetings) | 1.35 | 0.97 | 0.50 |
| Quote value opened (£40,000 each) | £54,000 | £38,610 | £19,980 |
| Orders (25 percent win rate) | 0.34 | 0.24 | 0.12 |
| Order value | £13,500 | £9,653 | £4,995 |
The arithmetic for the first column: 300 × 3 = 900 contacts. 900 × 2% = 18 replies. 18 ÷ 4 = 4.5 meetings. 4.5 × 30% = 1.35 opportunities, worth £54,000 in quotes. 1.35 × 25% = 0.34 orders, worth £13,500.
| Input | Type |
|---|---|
| 2 percent reply rate | KS Media planning target |
| 1.43 and 0.74 percent reply rates | Published benchmarks (Smartlead, per contact) |
| One meeting per four replies | KS Media planning assumption |
| 30 percent of meetings to opportunities, 25 percent win rate, £40,000 average order | Examples. Replace with your own |
Why isn’t one list enough?
Run the chain backwards for one order at the example rates. One order needs four opportunities. Four opportunities need about 13.3 meetings. That needs about 53 replies, which at 2 percent is about 2,667 contacts, or about 889 accounts at three contacts each. At Smartlead’s median it is about 2,402 accounts.
KS Media runs the same chain on its own sales, published on our homepage as an illustrative model: 300 accounts, 900 contacts, 18 replies, 4.5 calls, 1.35 Diagnostics at 30 percent of calls, and 0.54 Engine clients at 40 percent of Diagnostics. All of those rates are our planning assumptions. Read plainly, one 300-account list worked once by email alone is worth about half a client.
So the answer is to run outbound differently, in three ways:
- Choose accounts on a live signal. A firm that is hiring a sales engineer, exhibiting, under a new MD or launching something has a reason to reply now.
- Refresh the list every month. New signals appear all the time; a list chosen once goes stale.
- Add other sources of conversations. Referrals, LinkedIn, trade shows and technical content each bring in meetings the email list never would.
Where does the chain usually break?
Each rate tells you something different when it is low, and the fix is different at each stage.
| Weak rate | What it usually points to | What to change first |
|---|---|---|
| Few replies | The wrong accounts, the wrong people, or a message about you rather than them | Choose accounts on a live signal; open on their situation; shorten the first email |
| Replies, but few meetings | The ask is too big, or nothing in the email is worth a meeting | Offer something useful first: an application note, a short teardown, an answer to a question |
| Meetings, but few quotes | Poor fit, or meetings with people who can’t buy | Tighten the account list; check you are reaching the specifier or budget holder |
| Quotes, but few orders | Price, specification or timing, or a competitor already on the shortlist | Look at who wins and why; get in earlier, before the specification is written |
Change one stage at a time. If you rewrite the message and the account list in the same month, you won’t know which one moved the reply rate.
How long does it take?
Longer than a quarter. 6sense found the average B2B buying cycle was just over 10 months in 2025, and that buyers first spoke to a supplier about 61 percent of the way through their buying journey. Both figures cover all B2B sectors worldwide.
Source: 6sense, 2025 B2B Buyer Experience Report, nearly 4,000 buyers, global, checked 1 October 2026.
So a meeting held this month may become an order next year. Judge the work in the meantime on leading indicators: accounts engaged, meetings held, opportunities opened and quote value. Our method page defines each one.
What should you track every week?
| Measure | Why |
|---|---|
| Emails sent, by step | The volume the rates apply to |
| Bounces | Above 2 percent, stop and clean the list |
| Replies, per contact | Your reply rate, counted the same way as Smartlead’s |
| Positive replies | The share that wants to talk |
| Meetings held | Not booked; held |
| Opportunities opened | Quotes, tenders or framework applications with a value |
| Quote value | What the opportunities are worth |
Count replies per contact, not per email. It makes your rate comparable with Smartlead’s, and follow-ups can’t flatter it.
Where to start
Put your own rates into the pipeline calculator. It works forwards from an account list to orders, shows every line of the sum, and lets you set the reply rate to the 0.74 percent median to see the cautious case. If you want the rates pulled from your quote history and tested on a live segment, that is part of the Pipeline Diagnostic.
Quick answers
It depends on how it is counted. Smartlead’s published median is 0.74 percent of contacts, with the top quarter at 1.43 percent. Instantly’s 2025 average is 3.43 percent, but it divides replies by emails sent, not by contacts. Both count every reply, including ‘no thanks’. Compare your own rate with the benchmark that counts the same way, and never average the two.
Divide one by your win rate on opportunities, then by the share of meetings that become an opportunity. At a 25 percent win rate and 30 percent of meetings turning into a quote, one order takes about 13 meetings. Those two rates are examples; your own quote history gives the real ones.
There is no single set, because each stage is counted differently by different sources. For cold email, published reply rates run from Smartlead’s 0.74 percent median per contact to Instantly’s 3.43 percent average per email sent. For the later stages, meeting to quote and quote to order, your own history is a better benchmark than any published figure.
Define a lead first. Here it means a named person who has agreed to a meeting. Then work back from orders: orders needed divided by your win rate gives opportunities, and opportunities divided by the share of meetings that become a quote gives the meetings, which is the number of leads you need. Divide again by your meeting rate and reply rate to get the contacts to write to.
Sources
- Smartlead, Average cold email reply rate in 2026, checked 1 Oct 2026.
- Instantly, Cold Email Benchmark Report 2026 (data 1 January to 18 December 2025), checked 1 Oct 2026.
- 6sense, 2025 B2B Buyer Experience Report, checked 1 Oct 2026.
- KS Media, How we measure, checked 1 Oct 2026.