A CMO I worked with last year sent her CEO a slide that said marketing had generated 4,200 MQLs in Q2. Forty percent over plan. The CRO sent his own slide the same week. Sales had worked 980 of them and closed two deals. Two.
The CEO asked one question. "Are we selling to the wrong people, or are we calling the wrong people leads?"
Both. Both was the answer.
This is what sales and marketing alignment actually looks like before someone fixes it. Two scoreboards, two definitions of a lead, two narratives about what is working, and a pipeline that is mostly air. The leadership offsite version of alignment, where everyone agrees in a conference room and then goes back to optimising their own funnel, does nothing. I have been in the room for plenty of those offsites. None of them moved a number.
What moves the number is an SLA. A boring, written, measurable agreement that says exactly what marketing owes sales, exactly what sales owes marketing, and exactly what happens when either side breaks the contract. After ten years of this work, I think the SLA is the only mechanism that consistently fixes the alignment problem in B2B companies under 500 people. Everything else is therapy.
Alignment is not a meeting. It is a contract.
If you cannot write down what marketing owes sales this quarter and what sales owes marketing in return, the two teams are not aligned. They are coexisting.
Why the typical alignment fixes do nothing
Search "sales and marketing alignment" and you get the same five suggestions on every blog. Shared goals. Regular meetings. A common CRM. Joint planning. Better communication. All of it sounds right. None of it is actionable.
Shared goals fail because the two teams measure them differently. Marketing counts MQLs at the form fill. Sales counts pipeline at first meeting accepted. The same lead is a hit for marketing and a miss for sales, and at the end of the quarter both teams are technically right about their own number.
Regular meetings fail because they have no agenda that survives contact with a missed target. The weekly sync turns into a status update. Marketing says they hit their MQL number. Sales says the leads were bad. Marketing says the SDRs are not working them fast enough. Nobody changes anything. Next week, same script.
A common CRM fails because nobody trusts the data inside it. Marketing automation pushes MQLs into HubSpot or Salesforce. Sales reps mark them as bad without writing a reason. Marketing cannot tell whether the lead really was bad or whether the rep was busy with a hot inbound from a different source. Two months later, marketing is still optimising for the same junk MQLs.
I have seen all three failure modes at series A and B companies. They are not fixable with more meetings. They are fixable with a contract that defines exactly what each side has to deliver and what counts as a breach.
What an SLA actually looks like
A working sales and marketing SLA has four parts. Most teams write down one of them and call it a day. You need all four.
Part one: the definitions. What is a lead. What is an MQL. What is an SQL. What is a sales accepted lead. What is pipeline. Written down, with exact criteria, agreed by both VPs in writing. If your MQL definition is "anyone who filled out a form", you do not have a definition. You have a category.
Part two: the volume and quality commitments. How many MQLs marketing will deliver per quarter, broken down by segment. What percentage of those MQLs sales will accept within 24 hours. What conversion rate from MQL to SQL is acceptable, and what triggers a quality review.
Part three: the speed-to-lead commitment. How fast sales has to touch an MQL after it lands. How many touches in the first 14 days. What happens if a rep does not hit the cadence.
Part four: the reporting and review mechanism. Where the numbers live. Who sees them. How often the SLA gets reviewed. What the escalation path is when either side breaks the contract.
That is it. Four parts. If you write all four, get both leaders to sign it, and review it monthly with the data in the room, you have an SLA. If you skip any one, you have a wish list.
The SLA template I actually use
Here is the version I have rolled out at four companies in the past three years. Adjust the numbers to your business, but keep the structure.
Lead and stage definitions
Lead. Any record in the CRM with valid email and company name.
Marketing qualified lead. A lead that hits a score of 75 or higher, where score is built from firmographic fit (40 points), behavioural intent (35 points), and engagement recency (25 points). The full scoring model lives in the CRM, not in a spreadsheet.
Sales accepted lead. An MQL that an SDR has touched within 24 business hours and confirmed as worth working. The SDR has to choose one of three dispositions in the CRM. Accepted, recycled with reason, or rejected with reason. No silent ignoring.
Sales qualified lead. A sales accepted lead that has agreed to a discovery call. The first meeting is on the calendar.
Pipeline. A sales qualified lead that has reached stage two in the deal pipeline, where the customer has confirmed a problem we can solve and budget exists or can be built.
Marketing commitments
Marketing will deliver 1,200 MQLs per quarter, split 70 percent ICP segment A (sub-200-employee SaaS), 25 percent segment B (200-1000 employee SaaS), 5 percent other. At least 65 percent of MQLs must convert to SAL. If conversion drops below 60 percent for two consecutive months, marketing triggers an ICP and scoring review within 14 days.
Marketing will provide an enrichment record on every MQL inside HubSpot, with employee count, industry, tech stack, and last source. No MQL passes without an enrichment record.
Sales commitments
Sales will accept or recycle every MQL within 24 business hours. SDRs will run a 10-touch sequence over 14 days on accepted leads. If a lead is recycled, the SDR will write a recycle reason from a fixed list of 12 reasons. No free text.
If an SDR misses the 24-hour SLA on more than 5 percent of leads in a month, that SDR enters a one-week coaching plan with the SDR manager.
Closed-loop reporting
Both teams look at the same dashboard, weekly, in the same meeting. The dashboard shows MQL volume by segment, MQL to SAL conversion, MQL to SQL conversion, MQL to closed-won conversion, recycle reasons grouped by frequency, and time to first touch. The CRM is the source. The spreadsheet does not exist.
Escalation
If marketing misses MQL volume by more than 15 percent for a month, marketing presents a recovery plan within seven days. If sales misses SLA on speed-to-lead, the head of sales presents a recovery plan. If MQL to SAL conversion drops below threshold for two months running, both leaders meet with the CEO and walk through the data.
The numbers above come from the last three SLA rollouts I ran. The 38 percent lift is the median. The lowest was 22 percent, the highest was 71 percent. The 62 percent figure comes from a Forrester survey from late last year. Most teams still operate on handshake alignment.
How to roll out an SLA without a riot
Writing the document is the easy part. Getting two teams to live by it is where most rollouts fail. Here is the sequence I use.
The audit step is the one most teams skip. Do not skip it. Pull a sample of 200 MQLs from the past quarter. For each one, check whether sales touched it within 24 hours, whether the SDR disposition exists, and whether the lead actually fit the ICP. You will find that something like 40 to 60 percent of your MQLs are touched late, never dispositioned, or off-ICP. That data does the hard work of getting both leaders to admit there is a real problem.
I have done this audit at every company I have run alignment for. Every single time, the actual MQL to SQL conversion is between 20 and 35 percent lower than what the marketing dashboard shows. Sales does not trust the dashboard for a reason.
The closed-loop reporting layer
The SLA fails without honest reporting. Honest reporting fails without three things.
One source of truth. The CRM holds the data. Not the marketing automation tool. Not a spreadsheet that the ops manager updates every Friday. Both teams look at the same dashboard, sourced from the same record.
Recycle reasons as a fixed list. When an SDR rejects or recycles a lead, they pick from a dropdown of 10 to 15 options. Wrong title. Wrong company size. Not in territory. Already a customer. Already in pipeline. Did not respond after 10 touches. Each reason has a meaning. Free text turns into noise within a month.
Velocity reporting, not just conversion. MQL to SQL conversion matters, but so does the time between stages. If marketing is delivering MQLs that take 90 days to convert versus a benchmark of 30 days, the lead quality is bad regardless of what the conversion rate says.
The dashboard I build for every alignment rollout has six panels. MQL volume by source and segment. MQL to SAL conversion by segment with target line. Time to first touch by SDR. Recycle reasons ranked by volume. MQL to SQL conversion trend over 12 weeks. MQL to closed-won by source, lagged 90 days.
That is the meeting. Both teams in front of one dashboard. No slides. No narrative. Just the numbers and one question. What changes this week?
Tooling that actually carries the SLA
You do not need a big stack to run alignment. You need a CRM that holds the source of truth, a marketing automation layer that pushes MQLs in, an enrichment layer, and a reporting layer. Four things.
For automation between layers, I default to n8n on our own infrastructure for the data flows that matter to GDPR-sensitive work. For lighter glue, native CRM workflows are enough. The point is not the tool. The point is that the SLA mechanics are coded into the system. The recycle dropdown is required. The 24-hour SLA fires an alert. The dashboard is locked.
The deeper our CRM and RevOps build work goes, the more I have come to believe that alignment lives in the system, not in the strategy deck. A team with a clear SLA encoded in HubSpot will outperform a team with a beautiful alignment playbook in Notion every time.
The metrics that prove alignment is working
After 90 days of running an SLA, you should see five things move.
MQL to SQL conversion goes up. The median lift in my engagements is in the high 30s. If it is not moving, your scoring model is broken or your SDRs are not following the cadence.
Time to first touch drops. From a typical 18 to 36 hours down to under 6. This single number is responsible for half the conversion lift.
Recycle reasons concentrate. After two months, three or four reasons account for 70 percent of recycles. That is the signal for marketing on what to fix next.
Pipeline created from marketing-sourced leads grows. Not just MQL volume. Actual pipeline dollars. If MQL volume is up but pipeline is flat, the SLA is exposing a lead quality problem you can now solve.
Forecast accuracy improves. When both teams trust the same data, the forecast stops being a negotiation. I have seen forecast accuracy go from sub-70 percent to over 85 percent in two quarters, purely on the back of a working SLA.
Forecast accuracy I see at companies where the SLA has been live for two full quarters. Up from a sub-70 percent baseline at most series A and B teams.
The mistakes I see most often
The biggest mistake is writing the SLA without an audit. Both leaders agree on numbers they think are reasonable, the document gets signed, and the first month of data blows it up. Always audit first. Pull 200 leads. Walk the journey end to end. Build the SLA against reality, not against ambition.
The second mistake is making MQL volume the lead metric. If marketing is judged on MQL volume, they will hit MQL volume by lowering the bar. Always pair the volume target with a conversion target. The two numbers together protect both teams.
The third mistake is reviewing the SLA quarterly instead of weekly. Quarterly is too slow. By the time you see a problem, you have already wasted a quarter of effort. Weekly is the right cadence for the data. Quarterly is the right cadence for the SLA itself.
The fourth mistake is letting sales recycle leads with free-text reasons. Within a month, you have 400 unique recycle reasons. None of them are useful. Fix the dropdown. Audit it monthly.
The fifth mistake is treating the SLA as a marketing project. The SLA belongs to RevOps or to the COO. If marketing owns it, sales does not buy in. If sales owns it, marketing feels cornered. A neutral owner with budget across both is the only setup that lasts.
The work upstream of this, like getting your ideal customer profile right and shipping a lead scoring model that the SDR team trusts, is what gives the SLA something real to measure. Without those, you are writing a contract on a foundation that wobbles.
Want help drafting your SLA?
We run alignment audits and write SLAs for B2B teams that have outgrown the handshake version. Two-week sprints, both leaders signed by the end.
Book an audit →Frequently asked questions
How long does it take to roll out a sales and marketing SLA?
The document takes a week. The audit and the pilot take another four weeks. Full operating cadence with a working dashboard, six to eight weeks from kickoff. Anyone who tells you it is faster has not done one in a real company.
Does an SLA work for product-led companies?
Yes, with adjustments. Replace MQL with PQL (product qualified lead) as the marketing-to-sales handoff. Replace form-fill behavioural signals with in-product events. The four parts of the SLA are the same. The signals feeding the scoring model change.
What if our marketing team is too small to commit to MQL volume?
Then you do not need an SLA yet. You need a marketing automation foundation and a clear ICP first. SLAs work for teams shipping at least 200 MQLs per quarter. Below that, the law of small numbers makes the SLA look broken even when it is not.
Should the SLA cover customer success?
Eventually, yes. The first version is sales and marketing. After three months, add a CS handoff SLA that defines what sales owes CS at close and what CS owes sales for expansion signals. Trying to do all three from day one is too much change for most teams to absorb.
How do I get my CEO to sign off on this?
Show them the audit. Pull the 200-lead sample. Walk through the conversion math. Most CEOs sign off the moment they see how much pipeline is leaking between MQL and SAL. The SLA sells itself when the numbers are in front of you.
Wrapping up
Sales and marketing alignment is not a culture problem. It is a contract problem. When two teams have different definitions, different scoreboards, and no enforcement mechanism, they will drift, no matter how much they like each other. An SLA fixes drift by making the agreement specific, measurable, and reviewed often enough to matter.
If you are running a B2B company with a sales team of five or more and a marketing team that sends leads over the fence, write the SLA. Audit the data. Pilot one segment. Roll it out. The lift in conversion alone pays for the work three times over, and the side effect of having two teams pointed at the same number is worth more than any offsite ever was.
If you want help running the audit or writing the document, that is what we do at Ziel Lab. Two-week sprints, both leaders signed off, dashboard live by week three.