Vela RH · Monday report · marketing and demand generation view · data from Oct 5, 2026 · fictitious data
Where does Vela's revenue come from, and where is the waste?
Adjust the mix A cheap MQL isn't a good MQL.
Meta has the 2nd lowest cost per MQL (R$311) and a 0.0x return. LinkedIn ABM costs R$1,175 per MQL and returns 16.5x.
- Spend · Jul to Sep 2026R$163Kpaid media and events
- Paid media return4.2xARR won by the cohort ÷ spend
- Cost per paid MQLR$ 589212 paid media MQLs
- Session → lead1.7%64,930 sessions in GA4
Paid social converts 3.0% of sessions into leads; the site average is 1.7%Sessions and leads by channel (GA4)
Explain this chart
- What it shows
- Out of every 100 site visits, how many turned into a contact request, by the visit's source channel.
- Why measure it
- It's the first filter in the funnel and the only one before the CRM. It separates channels that bring visits from channels that bring interest, and points to pages that get traffic but don't convert.
- How to read it
- Longer bar = the channel that converts the most visits into leads. The text next to it shows the volume of sessions and leads, so a small channel isn't mistaken for a good one.
- Where the number comes from
- GA4: sessions and the form-submitted event (generate_lead) from July to September 2026 (Q3), in GA4's default channel grouping. There are no companies at this stage yet, so it doesn't open an account list.
Why it matters: It's the only number before the CRM. A channel that converts well here and poorly in the funnel brings in the curious, not buyers.
See the data
| Channel | Sessions | Change | Leads | Session → lead | Before |
|---|---|---|---|---|---|
| Organic search | 25,449 | +4.1% | 263 | 1.0% | 1.0% |
| Paid social | 17,676 | +13.4% | 532 | 3.0% | 3.0% |
| Paid search | 8,407 | +34.8% | 188 | 2.2% | 2.1% |
| Direct | 7,318 | -27.4% | 89 | 1.2% | 1.7% |
| Organic social | 2,449 | +1.9% | 5 | 0.2% | 0.2% |
| 2,019 | -1.6% | 11 | 0.5% | 0.7% | |
| Referral | 1,612 | +3.9% | 29 | 1.8% | 2.2% |
Paid social brings the most MQLs (122), and only 33.6% become SALsMQL → SAL by channel
Explain this chart
- What it shows
- What share of each channel's MQLs was accepted by the SDR (became an SAL).
- Why measure it
- It measures whether marketing brings leads the SDR considers good. A channel with high volume and low acceptance spends budget and pre-sales time on people who won't buy.
- How to read it
- The dashed line is the channel average. Below it, the channel sends leads the SDR doesn't accept. The highlight is the channel with the most volume.
- Where the number comes from
- CRM: companies that became MQLs between July and September (the Q3 cohort) and how many of them have reached SAL so far. Channels with fewer than 20 MQLs are left out of the chart.
- On click
- Click a channel to see its accounts: name, size, ICP, stage, MRR, SDR, rep and the cost per MQL of the source campaign.
Why it matters: The channel average is 49.2%. An MQL the SDR doesn't accept costs budget and pre-sales time.
See the data
| Channel | MQL | SAL | SQL | Proposal | Won | MQL → SAL | Proposal → won | MRR won |
|---|---|---|---|---|---|---|---|---|
| Paid social | 122 | 41 | 17 | 8 | 2 | 33.6% | 25.0% | R$37K |
| Organic search | 96 | 58 | 38 | 26 | 0 | 60.4% | 0.0% | R$ 0 |
| Paid search | 90 | 52 | 44 | 25 | 1 | 57.8% | 4.0% | R$ 6,580 |
| Event | 46 | 24 | 5 | 0 | 0 | 52.2% | n/d | R$ 0 |
| Outbound (SDR) | 0 | 43 | 18 | 13 | 0 | n/d | 0.0% | R$ 0 |
| Direct | 42 | 20 | 18 | 11 | 0 | 47.6% | 0.0% | R$ 0 |
| Referral | 17 | 11 | 7 | 3 | 1 | 64.7% | 33.3% | R$ 4,180 |
| 3 | 1 | 1 | 1 | 0 | 33.3% | 0.0% | R$ 0 |
The cheapest campaign per MQL isn't the one that brings revenueCampaigns: cost per MQL and return
- Google · brand41 MQLs · 1 win · R$9,048 R$ 221per MQL 8.7x
- Meta · SMB leads88 MQLs · 0 wins · R$27K R$ 311per MQL 0.0x
- Event · HR trade show46 MQLs · 0 wins · R$38K R$ 826per MQL 0.0x
- Meta · remarketing5 MQLs · 0 wins · R$4,508 R$ 902per MQL 0.0x
- Google · payroll and time tracking search36 MQLs · 0 wins · R$33K R$ 917per MQL 0.0x
- Google · competitors13 MQLs · 0 wins · R$12K R$ 921per MQL 0.0x
- LinkedIn · ABM enterprise23 MQLs · 2 wins · R$27K R$ 1,175per MQL 16.5x
- LinkedIn · payroll e-book6 MQLs · 0 wins · R$12K R$ 1,996per MQL 0.0x
Explain this chart
- What it shows
- For each paid campaign, the cost per MQL and the return: how much annual revenue its cohort has already brought for every real invested.
- Why measure it
- Cost per MQL is easy to optimize and misleading: it rewards cheap leads. Cohort return shows which campaign pays the bill, and that's what should decide where the budget goes.
- How to read it
- The buttons reorder the list: cheapest per MQL or highest return. Green bar = return above 1x (it paid for itself); coral = below.
- Where the number comes from
- Paid media (Google, Meta, LinkedIn) and the event cost, from July to September, divided by the cohort accounts that reached each stage. Return = ARR won by the cohort ÷ spend.
- On click
- Click a campaign to see its accounts, with each one's cost per MQL.
Why it matters: Optimizing for cost per MQL rewards volume. What pays the bills is return by cohort. The quarter's cohort is still maturing: every campaign's return tends to go up, but the comparison between them already holds.
See the data
| Campaign | Spend | MQL | Por MQL | Por SAL | Por SQL | Per proposal | Wins | ARR won | Return |
|---|---|---|---|---|---|---|---|---|---|
| Event · HR trade show | R$38K | 46 | R$ 826 | R$ 1,583 | R$ 7,600 | n/d | 0 | R$ 0 | 0.0x |
| Google · payroll and time tracking search | R$33K | 36 | R$ 917 | R$ 1,834 | R$ 1,942 | R$ 3,301 | 0 | R$ 0 | 0.0x |
| Meta · SMB leads | R$27K | 88 | R$ 311 | R$ 1,245 | R$ 3,042 | R$ 6,845 | 0 | R$ 0 | 0.0x |
| LinkedIn · ABM enterprise | R$27K | 23 | R$ 1,175 | R$ 2,253 | R$ 5,407 | R$ 6,759 | 2 | R$446K | 16.5x |
| LinkedIn · payroll e-book | R$12K | 6 | R$ 1,996 | R$ 5,988 | R$12K | n/d | 0 | R$ 0 | 0.0x |
| Google · competitors | R$12K | 13 | R$ 921 | R$ 1,496 | R$ 2,394 | R$ 3,990 | 0 | R$ 0 | 0.0x |
| Google · brand | R$ 9,048 | 41 | R$ 221 | R$ 348 | R$ 411 | R$ 754 | 1 | R$79K | 8.7x |
| Meta · remarketing | R$ 4,508 | 5 | R$ 902 | R$ 902 | R$ 2,254 | n/d | 0 | R$ 0 | 0.0x |
Paid social brings volume outside the profile: 49.2% in the ICPICP accounts by channel
Explain this chart
- What it shows
- What share of each channel's accounts fits the ideal customer profile (mid-sized companies and up).
- Why measure it
- Off-profile leads don't become customers, or become customers who leave early. The ICP share by channel shows where marketing is buying volume instead of opportunity.
- How to read it
- The longer the bar, the more the channel brings in people who can buy. In coral, the channel with the smallest share in the ICP.
- Where the number comes from
- CRM: company size for the Q3 cohort (July to September). ICP = Mid-market (200 to 999 employees), Enterprise and Large Enterprise.
- On click
- Click a channel to see its accounts: name, size, ICP, stage, MRR, SDR, rep and the cost per MQL of the source campaign.
Why it matters: In the other channels, most accounts are in the ICP. It's the same reason behind Meta's low return.
Of the accounts that came in during Q3/2026, inbound brought 4 of the 4 wins so farWins by source
Explain this chart
- What it shows
- How many contracts came from inbound (the lead came looking for Vela), outbound (the SDR went looking for the lead) and events.
- Why measure it
- It shows where the contract comes from, not just the lead. It tells you whether outbound or the event is worth keeping, by comparing what each source closes with what it costs.
- How to read it
- Longer bar = the source that closes the most. The text next to it shows accounts, MRR won and the source's media spend.
- Where the number comes from
- CRM: contracts won by accounts in the Q3 cohort (July to September), grouped by the deal's source.
- On click
- Click a source to see the accounts won, with the source campaign's cost per win.
Why it matters: R$48K of new MRR came from inbound. The 16 contracts closed in the quarter (executive view) mostly come from accounts that came in earlier: this cohort is still maturing. Outbound isn't a revenue channel yet.
See the data
| Source | Accounts | SAL | SQL | Proposal | Won | MRR won | Spend |
|---|---|---|---|---|---|---|---|
| Inbound | 370 | 183 | 125 | 74 | 4 | R$48K | R$125K |
| Event | 46 | 24 | 5 | 0 | 0 | R$ 0 | R$38K |
| Outbound | 43 | 43 | 18 | 13 | 0 | R$ 0 | R$ 0 |
The event cost R$38K and didn't reach a proposalEvent funnel
- 46MQL
- 24SAL
- 5SQL
- 0Proposal
- 0Won
Explain this chart
- What it shows
- How far the accounts from the HR trade show got: MQL, SAL, SQL, proposal and won.
- Why measure it
- Events are expensive and hard to measure: without following the accounts through to proposal, all you have is the feeling that it went well. The funnel shows whether it generated real pipeline.
- How to read it
- Each bar is a stage, sized by the number of accounts that reached it. Coral = no account got there.
- Where the number comes from
- CRM: accounts whose source campaign was the event, in the Q3 cohort (July to September). The event cost counts as the campaign's spend.
- On click
- Click a stage to see the accounts that reached it, with the event cost per account at that stage (per MQL, per SAL, per SQL or per opportunity).
Why it matters: It generated conversations, not pipeline. Before the next edition, define how the event's accounts turn into meetings.
Before deciding: what can distort these numbers
- Between Jun 8 and Jun 21, 2026, 56% of new contacts arrived as DIRECT_TRAFFIC, against 8% in the previous weeks. It could be a broken UTM: paid leads without attribution inflate Direct and disappear from the campaigns' numbers.
- google_ads had no spend for 15 days (May 12 to May 26, 2026). Leads and MQLs from that platform drop in the period because of the pause, not performance.
What to do this week
- Pull budget from Meta for SMB leads (0.0x) and test more LinkedIn ABM (16.5x).
- A company size filter on the paid social forms: today 49.2% of the accounts are in the ICP.
- Pause the payroll e-book: the most expensive MQL, with no wins.
- Read the comparison with the previous quarter discounting the Google Ads pause (May) and the UTM break (June): both pulled down the comparison base.