Gemini 4 Carbon: What Ledgerline's accountancy practices want built next

Scorecard

The analysis

Findings

1. Answer to the CTO: Do not tell the board the survey backs building AI categorisation next. Its first-place plurality is a statistical mirage caused by split-voting among multi-person paid firms, sample over-representation of heavy daily users, and sharp polarisation: AI categorisation is actually ranked dead last (4th) by more respondents than any other option, and by two-thirds of paying practices.

   High confidenceBase 548

   Unweighted, AI categorisation wins 37.8% of first-place votes (n=548) vs 34.9% for reliable bank feeds, 15.7% for bulk MTD, and 11.7% for client portal. However, 42.2% of all respondents and 65.5% of paid practices (n=296) rank AI categorisation 4th (lowest priority), giving it a worse overall mean rank (2.57) than reliable bank feeds (1.99; paid mean 1.76). Support comes overwhelmingly from solo free-tier users (70.2% of free plan rank it 1st, n=252; 60.9% of solo practices, n=258) vs just 10.1% of paid practices (n=296). Even reweighting by plan tier alone (85% free, 15% paid) gives reliable bank feeds a better population mean rank (2.18) than bank feeds' nearest rivals when accounting for paid revenue weight—where 100% of current revenue sits.

2. More reliable bank feeds is the clear consensus priority across both paying customers and the broader user base, and it relieves a major root cause of manual coding complaints.

   High confidenceBase 548

   Among paid practices (n=296, representing 100% of revenue), 48.3% rank reliable bank feeds 1st and only 5.1% rank it last (mean rank 1.76). Across all respondents (n=548), 73.7% rank reliable bank feeds 1st or 2nd, and only 7.5% rank it last (mean rank 1.99). In open-text responses (n=186), broken, disconnected, missing, or delayed bank feeds are the #1 complaint (69 mentions, 37.1%), with 22 respondents explicitly stating that manual coding surges because feeds drop out, expire every 90 days, or arrive late.

3. Answer to the Finance Director: Do not bank on £8/user/month add-on revenue from AI categorisation based on the raw 27.7% 'Yes' figure. Paying firms largely reject it, free solo users face a base-plan paywall hurdle, and stated willingness-to-pay overstates actual purchase.

   Medium confidenceBase 548

   Only 16.9% of paid practices (50 of 296) say 'Yes' to paying £8/user/month (57.4% say 'No'), and willingness to pay drops as firm size (and seat count) grows: 36.8% for 'Just me' (n=258), 21.5% for 2–9 people (n=172), 17.0% for 10–19 people (n=106), and 16.7% for 20+ people (n=12). Even if paid respondents matched all 1,230 paid practices at stated face value (~208 practices) and averaged ~4–6 paid seats, unadjusted stated intent among paid accounts implies roughly £80k–£120k/year—and realistic B2B conversion of stated 'Yes' (typically 25%–50%) cuts that to ~£20k–£60k/year from the paid base. While 40.5% of free practices (n=252, all 1-user capped) say 'Yes'—theoretically ~2,820 of 6,970 free accounts (£271k/year if sold standalone at £8/mo)—the survey did not test whether free users must first upgrade to a paid base plan, nor adjust for heavy-user sample skew (45.2% of free respondents use Ledgerline 5+ days/week vs 18% of all accounts).

4. Answer to the Head of Sales: The survey cannot bear the weight of prioritising the client portal to win larger firms. The 100% figure for 20+ person firms rests on just 12 existing customers, while 10–19 person firms strongly prefer reliable bank feeds and bulk MTD.

   Low confidenceBase 118

   All 12 respondents in practices of 20+ people ranked the client portal 1st (100.0%, 95% Wilson CI ~76%–100%) and 4 of those 20+ practices mentioned chasing client paperwork in Q1 (L215, L509, L540, L544). However, n=12 is a tiny base of existing accounts (zero non-customer prospects were surveyed), and adjacent mid-sized practices of 10–19 people (n=106) ranked the client portal dead last among first-place choices at just 6.6% (27.4% ranked it 4th), heavily favouring reliable bank feeds (49.1% 1st) and bulk MTD submissions (30.2% 1st).

5. Bulk Making Tax Digital (MTD) submissions is the clear second-highest priority for revenue-generating practices, creating acute quarter-end bottlenecks as client counts scale.

   High confidenceBase 296

   26.4% of paid practices rank bulk MTD 1st (mean rank 2.23, second only to bank feeds at 1.76; only 7.8% rank it last), rising from 5.8% in solo practices to 22.7% in 2–9 person practices and 30.2% in 10–19 person practices. Extrapolated directly to 1,230 paid practices, ~325 paid firms rank it #1. Open-text responses contain 27 mentions of tedious one-by-one MTD filing eating entire days around the 7th of the month and VAT quarter-ends.

6. Unprompted open-text answers reveal an unmeasured fifth problem—application slowness and page/report load latency—that rivals bulk MTD and exceeds client document chasing.

   Medium confidenceBase 186

   27 of 186 Q1 respondents (14.5%, spanning free and paid plans across solo to 10–19 person firms) complained of slow page loads, sluggish search, lagging deadline views, and report timeouts—compared with 27 mentions of individual MTD filing and just 14 mentions (9 paid) of chasing client documents. Because app speed was omitted from Q2, its priority relative to feature work is unquantified.

Open-text themes it coded

Broken, disconnected, missing, or delayed bank feeds 74Manual categorisation and coding of bank transactions 59Lack of bulk Making Tax Digital (MTD) / VAT submissions 25Slow application performance, page loads, search, and reports 26Chasing clients for missing documents, statements, and receipts 15Nothing slows the practice down / satisfied with Ledgerline 25

The survey it planned

3 screening questions and 8 questions, as the model wrote them.

1. S1

   Is your accountancy or bookkeeping practice based in the UK?

   One answer
   • Yes
   • No

   Continues if Yes

2. S2

   Does your practice currently use Ledgerline?

   One answer
   • Yes
   • No

   Continues if Yes

3. S3

   Which of the following best describes your involvement in choosing or paying for software for your practice?

   One answer
   • I make the final decision on software for the practice
   • I share or influence decisions on software for the practice
   • I have no involvement in software decisions

   Continues if I make the final decision on software for the practice; I share or influence decisions on software for the practice

4. Q1

   Including yourself, how many people work in your practice (full-time or part-time)?

   One answer
   • 1 (sole practitioner, no other staff)
   • 2 to 9 people
   • 10 or more people
5. Q2

   Which Ledgerline plan is your practice currently on?

   One answer
   • Free plan (1 user)
   • Paid plan (charged per user per month)
   • Not sure
6. Q3

   In a typical working week, how often do you or members of your practice use Ledgerline?

   One answer
   • Every working day
   • 2 to 4 days a week
   • About once a week
   • Less than once a week
7. Q4

   Over the past 6 months, which of the following have slowed your practice down when using Ledgerline? Select all that apply.

   Any that apply
   • Broken bank-feed connections or having to reconnect bank feeds
   • Categorising bank transactions manually
   • Chasing clients for documents or records outside Ledgerline
   • Submitting Making Tax Digital (MTD) returns one client at a time
   • Updating client records or tracking deadlines
   • Slow loading speeds or general technical errors
   • Something else in Ledgerline
   • None of these — Ledgerline has not slowed us down
8. Q5

   In a sentence or two, what is the single biggest thing that slows your practice down in Ledgerline today, and how does it affect your work?

   Open text

   Routing Shown only if Q4 is: Broken bank-feed connections or having to reconnect bank feeds; Categorising bank transactions manually; Chasing clients for documents or records outside Ledgerline; Submitting Making Tax Digital (MTD) returns one client at a time; Updating client records or tracking deadlines; Slow loading speeds or general technical errors; Something else in Ledgerline

9. Q6

   We have room to build one major improvement to Ledgerline over the next six months. Please rank these four options from 1 (most valuable to your practice) to 4 (least valuable to your practice).

   Ranking
   • AI categorisation of bank transactions (automatically suggests categories for bank lines)
   • More reliable bank feeds (fewer broken connections and reconnections)
   • A client portal for requesting and receiving documents from clients
   • Bulk Making Tax Digital (MTD) submissions across multiple clients at once
10. Q7

    Suppose AI categorisation of bank transactions were offered as an optional paid add-on (charged on top of your current free or paid plan), rather than included at no extra cost. What would your practice most likely do?

    One answer
    • We would very likely pay extra for it as an add-on
    • We might consider paying extra, depending on the price
    • We would only use it if it were included in our plan at no extra cost
    • We would not use it, even at no extra cost
11. Q8

    What is the highest extra fee per user per month (excluding VAT) your practice would realistically pay for an AI bank-transaction categorisation add-on?

    One answer
    • Up to £5 per user per month
    • £6 to £10 per user per month
    • £11 to £20 per user per month
    • More than £20 per user per month
    • On reflection, we would not pay an extra monthly fee

    Routing Shown only if Q7 is: We would very likely pay extra for it as an add-on; We might consider paying extra, depending on the price