The Short Version
The AK035 CV Tailor tool grew fast between July and September 2026. It more than doubled.
But growth is not the whole story. One user told us something we cannot ignore. Their CV was too long for a recruiter to read. That comment matters more than any number in this report.
Here is what happened, what we learned, and what we are doing next.
What the Tools Do
Before the numbers, here is a quick reminder.
The AK035 platform offers three free tools:
- CV Tool – Build, improve, or tailor your CV for a specific job
- Career Guidance Tool – Answer questions about your career path
- Cover Letter and Email Drafting Tool – Write professional letters and emails
All three are free. No hidden costs. No expensive services. Just help.
The Big Change: Tailoring Beat Building
In Q2 2026, most people used the CV tool to review what they already had.
In Q3 2026, that flipped. People now want to tailor their CV for a specific job.
| Tool Action | Q2 Share | Q3 Share |
|---|---|---|
| Tailor | 18.7% | 46.4% |
| Review | 40.3% | 24.7% |
| Improve | 27.6% | 27.6% |
| Create | 1.3% | 1.3% |
Tailor use grew 169.7% per day. That means more people are taking a CV they already have and aiming it at one job.
The tool that builds a CV from scratch? Almost nobody uses it. Create dropped from 19 uses to 8.
What this tells us: People arrive with a CV. They do not want a new one. They want the one they have to work harder.
Standout Problem: Too Many Words
Here is where the report gets uncomfortable.
Backend work grew 24.4% per day. Frontend use grew 9.0% per day. That means each visit now produces more text than before.
Then a user wrote this:
“It’s a lot. Like the CV summary is long, the key skills, etc. Most recruiters prefer a CV that’s straight to the point and easy to read. A CV they can scan not a CV that takes more than 3min to read through.”
That user rated the tool 3 stars. They described their own reading experience, not a guess about what recruiters want.
We treat this as an open defect, not a preference.
If a recruiter cannot scan your CV in three minutes, the CV fails. It does not matter how good the words are.
Satisfaction: Mostly Good, But the Floor Fell
18 people left feedback between July and August. September recorded none.
| Rating | Entries |
|---|---|
| 5 stars | 12 |
| 4 stars | 3 |
| 3 stars | 1 |
| 2 stars | 1 |
| 1 star | 1 |
Mean rating: 4.33 out of 5. That is down from 4.69 in Q2.
The 2-star and 1-star ratings arrived on the same day in July. They came from two different sessions. No words explained what went wrong.
What this means: 16 sessions rated out of 291 that logged actions. The 4.33 describes people motivated enough to rate, not everyone who used the tool.
Median satisfaction stayed at 5 stars. Most raters remain happy. But we cannot fix what we cannot see.
What users wrote. Five of 18 entries carried text.
“Great tool which improved my cv” – improve, 5 stars, Bojanala Platinum
“I loved it” – improve, 5 stars, Mangaung
“I thought you where going to make me a new CV” – review, 5 stars, Mangaung
“Very good” – improve, 4 stars, City of Johannesburg
“It’s a lot. Like the CV summary is long, the key skills, etc. Most recruiters prefer a CV that’s straight to the point and easy to read. A CV they can scan not a CV that takes more than 3min to read through” – improve, 3 stars, City of Ekurhuleni
What this supports. Four of the five comments praise the tools. One names a defect. Written comment rates rose in proportion, from 4 of 35 entries in Q2 to 5 of 18, which is 27.8 per cent against 11.4 per cent.
What this cannot support. 16 sessions rated out of 291 sessions that logged actions, a response rate of 5.5 per cent. The 4.33 mean describes people motivated enough to rate, not the audience. The 2 star and 1 star entries carried no words, so the data cannot say what went wrong on 29 July.
Who Is Using the Platform
Where They Come From
| District | Q3 Actions | Q2 Actions |
|---|---|---|
| City of Tshwane | 117 | 71 |
| City of Johannesburg | 93 | 121 |
| Unknown | 90 | 14 |
| City of Ekurhuleni | 51 | 60 |
| Outside South Africa | 46 | 2 |
Tshwane took the lead from Johannesburg. It also carried the heaviest tailoring load, with 73 of its 117 actions being tailor requests.
Gauteng municipalities held 46.9% of all activity. KwaZulu-Natal held 12.5%. Western Cape held 7.4%.
One problem: 90 rows carried the label “Unknown.” That is up from 14 in Q2. We cannot say where those users came from. That limits what we can learn.
Career Guidance: Small but Rising
The Career Guidance tool ran 29 times across 22 days in Q3. That is tiny compared to 608 CV actions.
But September reached 14 requests. July had 7. That is a doubling, though from a very small base.
31 people completed a career guidance profile. Here is who they are:
- Age: Youngest 21, oldest 43, mean 28.7
- Education: 41.9% hold a degree or higher
- Location: 51.6% from Gauteng, 25.8% from KwaZulu-Natal
Location
| Province | Users | Share |
|---|---|---|
| Gauteng | 16 | 51.6% |
| KwaZulu-Natal | 8 | 25.8% |
| Free State | 2 | 6.5% |
| North West | 2 | 6.5% |
| Limpopo | 2 | 6.5% |
| Western Cape | 1 | 3.2% |
Education
| Qualification | Users |
|---|---|
| Diploma | 11 |
| Bachelor’s Degree | 8 |
| Grade 12 or Matric | 5 |
| Honours Degree | 5 |
| Certificate | 2 |
13 of 31 users, or 41.9 per cent, hold a degree or higher, against 33.3 per cent in Q2. Nobody holds a Master’s degree this quarter, where Q2 had one.
What They Told Us
| Problem | Users | Share |
|---|---|---|
| No responses from employers | 18 | 58.1% |
| Career guidance | 7 | 22.6% |
| Finding jobs | 4 | 12.9% |
| Upskilling | 2 | 6.5% |
Roughly six in ten hold a qualification, apply, and hear nothing back. That figure held steady across two quarters.
How Long They Have Searched
| Duration | Users |
|---|---|
| More than 2 years | 10 |
| 1 to 2 years | 7 |
| 3 to 6 months | 5 |
| 6 to 12 months | 4 |
| Less than 3 months | 4 |
17 of 31 users, or 54.8%, have searched longer than a year. In Q2, that figure was 37.5%. People arriving now have been knocking on doors longer than the group before them.
What They Will Accept
- Office-based: 80.6%
- Willing to relocate: 80.6%
- Remote work: 71.0%
- Hybrid work: 67.7%
- Willing to start a business: 22.6%
Eight in ten would move province for work. Locality does not protect a job seeker here.
The Traffic Story: One Week Carried September
The daily traffic export covers 4 September to 30 September. July and August produced no daily records.
What 27 days showed:
- Page views: 1,423
- Impressions: 3,029
- Clicks: 109
Five days from 14 to 18 September carried 633 page views. That is 44.5% of the September window in five days.
The last seven days produced 269 page views. The first seven produced 321.
The spike was one event, not lasting growth.
Devices: Phones Rule, But Ads Suffer
The device export says 79.3% of page views come from high-end mobile phones. Desktop takes 20.1%.
Phone users produce 94.3% of clicks. But phone users see fewer ads than desktop users.
- Mobile ad visibility: 56.0%
- Desktop ad visibility: 71.4%
Where an ad sits on a small screen decides whether a free tool stays free.
The device export disagrees with September’s daily rows by 3.04 times. We cannot resolve that conflict, so both figures appear as given.
Ten questions the quarter answers
- Do users want to build a CV or improve one they already have? They want an existing CV aimed at one job. Tailor led at 282 actions and grew 169.7 per cent per day. Create, the tool that builds from nothing, fell to 8 uses.
- Is the tailor rise real, or a tracking failure? Real. Inside named districts, tailor held 37.5 to 37.9 per cent of activity between July and August while its count grew from 72 to 106. Capture failed on 90 rows, and those failures do not explain the rise.
- Which district leads the platform now? City of Tshwane, at 117 actions against Johannesburg’s 93. Tshwane also runs the heaviest tailor mix, 73 of its 117 actions.
- Are users satisfied? Mostly, on thin evidence. 12 of 18 ratings were 5 stars. Two low scores on one July day pulled the mean to 4.33. 16 sessions rated out of 291.
- Is the career guidance tool being used? Barely. It ran 29 times across 22 days, against 608 CV actions. September doubled July’s monthly rate from a small base.
- Can 31 profile completions speak for the user base? No. These people chose to fill in a long form. They hold more degrees than the national average, and they cluster in Gauteng. They describe the most motivated end of the audience.
- What frustrates users most? Silence. 18 of 31 named no responses from employers. That figure held near six in ten for two quarters.
- Are people finding work quickly? No, and it worsened. 17 of 31 have searched longer than a year, against 37.5 per cent in Q2. Ten passed the two-year mark.
- Did the September traffic spike mean lasting growth? No sign of it. Five days from 14 to 18 September carried 633 page views, 44.5 per cent of the September window. The last seven days then produced 269 page views against 321 in the first seven. The spike was one event, and the window opened on it.
- Do users want to move for work? Yes. 25 of 31 would relocate. Q2 recorded 79.2 per cent willing to relocate, and the appetite held while the search dragged on.
What We Are Doing Next
We set seven priorities for Q4 2026. Here are the ones that matter most to you.
1. Capture the District on Every Tailor Request
Target: Cut unknown-district rows from 14.8% to under 5% by 31 December 2026.
Why: We cannot improve what we cannot measure.
2. Measure the Length of Every CV
Target: A length figure recorded for every tailor and improve request by 30 November 2026.
Why: The 3-star comment and the 24.4% rise in tokens per use make this the clearest open defect. We need to know how long our CVs are before we can shorten them.
3. Test a Shorter CV Output
Target: Lift the mean rating above 4.50 with at least 30 entries in the quarter.
Why: A CV that takes more than three minutes to read fails the user. We will offer a two-page output as the default.
4. Ask Unhappy Users What Went Wrong
Target: A written reason attached to every rating of 3 or lower by 31 December 2026.
Why: Two low scores on one day with no words cannot be fixed.
5. Offer Career Guidance to Repeat Tailor Users
Target: Lift guidance runs from 29 to 75 in the quarter.
Why: Someone reshaping one CV across 12 attempts needs direction, not another rewrite.
The Bottom Line
The CV Tailor tool grew because people want their existing CV aimed at a specific job. That is clear.
But growth exposed a problem. The tool produces too many words. One user said it plainly. A recruiter cannot scan a CV that takes more than three minutes to read.
We are fixing that. We are measuring CV length. We are testing shorter outputs. We are asking unhappy users what went wrong.
Information should be free, simple, and actionable. That is the AK035 belief. A CV that nobody reads is none of those things.
Try the Tools
Build a CV for free, work through the career guidance tool, or draft a cover letter at ak035.co.za
Read the other quarterly reports at ak035.co.za/reports
References
Statistics South Africa (2026) Quarterly Labour Force Survey, Quarter 2: 2026. Statistical release P0211. Pretoria: Stats SA. Published 11 August 2026.
South African Government News Agency (2026) SA official unemployment rate rises to 33.6%. Pretoria: SAnews.gov.za.