Fractional bookkeepers for Indian e-commerce: when software stops being enough

There is an inflection point in every Indian e-commerce business where software stops being enough. The signal is rarely dramatic — it is the third month in a row where the founder spends a Sunday afternoon untangling a Flipkart settlement variance, or the GST notice that arrives quoting Section 73 with a thirty-day response window, or the moment the audit clean-up is being done by an external CA who sends back a 200-row pivot table marked 'please reconcile'. Software has automated the matching; what remains is judgement, and judgement is what a fractional bookkeeper sells. This post is a decision framework for when an Indian e-commerce founder should hire a fractional bookkeeper, what to look for, and where the hand-off between tooling and human judgement actually sits in practice.

A fractional bookkeeper is not a junior accountant on retainer. The honest version of the role is a credentialled practitioner — usually an ICAI-registered CA, a registered GST practitioner, or an experienced practitioner working under a CA's supervision — who handles bookkeeping, GST filings, marketplace reconciliation, monthly close, and the long tail of judgement-bound activities that software does not own. Day-to-day, that means: filing GSTR-1 and GSTR-3B every month with ITC reconciliation against GSTR-2B JSON, responding to GST departmental notices when they arrive, reviewing marketplace settlement variances and filing disputes inside the carrier's window, doing TDS reconciliation against Form 26AS, preparing the monthly close pack with accruals and prior-period adjustments, and producing a founder-readable MIS dashboard with cash position, working capital, and category-level margin. Some practitioners also do day-to-day vendor and payroll posting in Tally, Zoho Books, or QuickBooks; others stay strictly at the review-and-close layer.

There are five thresholds at which the calculus shifts in favour of hiring a fractional bookkeeper rather than continuing with founder-time plus software. The first is order volume: below about 500 marketplace orders per month, a founder with a clean spreadsheet workflow can usually keep up with reconciliation review and dispute filing. Above 1,500 orders per month across two or more marketplaces, the per-row review work generally exceeds what any non-finance founder will sustain reliably. The second is marketplace count: a single-marketplace seller has one fee structure to learn; a three-marketplace seller has Amazon Settlement Report V2, Flipkart's four-fee structure, and Meesho's price-base nuance to track simultaneously, each with its own dispute window. The third is GST registration count: a single-state registration is straightforward; multi-state GSTINs with inter-state e-way bills, branch transfers, and place-of-supply rules become a notice-prone configuration that benefits materially from practitioner oversight.

The fourth threshold is dispute load. A business filing fewer than five marketplace disputes a month can document and submit them ad-hoc; a business filing twenty disputes a month with mixed weight-correction, commission-variance, and return-window claims needs a practitioner who tracks dispute outcomes systematically and adjusts evidence packets to lift win rates over time. The fifth is audit cycle. Once an external auditor is in the picture annually — for statutory audit, internal audit, or due-diligence prep ahead of a fundraise — the cost of having a fractional bookkeeper maintain audit-ready books month-to-month is materially less than the cost of a year-end clean-up sprint. The clean-up sprint also tends to surface errors that have been compounding for months, which the founder discovers under deadline pressure.

On cost, the trade-off is cleaner than most founders assume. Hourly rates in India for vetted fractional bookkeepers typically range from ₹500 to ₹2,500 per hour depending on credentials and specialty. A typical SMB retained engagement runs 8 to 20 hours a month, putting the monthly retainer commonly between ₹10,000 and ₹50,000. The comparison point is a full-time controller — ₹8 lakh to ₹15 lakh per year fully loaded — which a business doing ₹1–5 crore in annual revenue rarely needs at full-time scope. The intermediate option, an internal junior accountant on ₹25,000 a month, often ends up being the worst of both worlds: cheaper than a fractional but without credentials, supervisor coverage, or the ability to sign off on filings. The fractional model exists because most Indian SMBs need credential-bearing judgement on a part-time basis, and the credentials are what carry the risk on filings.

Where exactly does software hand off to a person? The clean way to think about it is that software handles the deterministic and probabilistic layer — file format detection, column mapping, matching settlements to orders, flagging variance, surfacing exceptions in a queue — and a fractional bookkeeper handles everything that requires judgement, accountability, or a credential. Concretely: dispositioning unmatched marketplace settlements where the variance is in a grey zone (₹2.17 difference, is it a fee or a duplicate?); responding to a GST departmental notice with the correct legal framing; deciding the threshold above which a write-off requires controller sign-off; choosing whether to file a dispute or absorb the variance based on probability of recovery and dispute-filing cost; signing off on the monthly close pack and the year-end reconciliation. None of these tasks belong on the matching path — they belong to a person with the credentials to defend the decision.

On credentials, the practical question is what to verify. ICAI membership confirms the practitioner has cleared the CA exams and is in good standing — verifiable by the membership number against the ICAI public registry. A GST practitioner ID confirms the practitioner is enrolled to file GST returns on behalf of taxpayers — verifiable on the GSTN portal. An experienced practitioner without either credential can still do excellent operational bookkeeping, but cannot independently sign off on statutory filings; the workaround in many small practices is supervision by a CA who counter-signs filings. For an Indian e-commerce business with even modest GST complexity, the safer default is to engage a CA or GSTP directly rather than rely on supervision arrangements that can break down if the supervising CA exits the relationship. The ReconPe directory lists the credential type on every profile and verifies it manually before issuing the Verified badge.

Fit evaluation is the part most founders underweight. Stack proficiency matters: a bookkeeper who has spent five years in a manufacturing-heavy Tally environment will need ramp time to be useful on a Zoho-Books-plus-Razorpay e-commerce setup; a Zoho Books native will be productive on day one. Marketplace experience matters more than generic e-commerce experience: an Amazon-only practitioner may not know the Flipkart four-fee structure or Meesho's price-base nuance. Response-time expectations should be agreed upfront — vetted fractionals typically respond by email within a business day; if your business has a 30-day GST notice cycle and the bookkeeper takes a week to reply, the engagement will not work even if the practitioner is technically excellent. The 30-minute call before any engagement starts is the most useful filter; ask for two recent marketplace reconciliation samples and one GST notice response, and read them.

Engagement model is the last decision. Hourly suits one-off work — an audit prep sprint, a GSTR-9 annual filing, a backlog clean-up — billed against tracked hours. Project suits a defined deliverable with a known scope, quoted upfront; this is the right shape for ad-hoc filings or one-time clean-ups. Retained suits ongoing work — monthly close, periodic GST filings, marketplace reconciliation review, MIS — billed at a discounted blended rate against a monthly hours allocation. Most growing Indian e-commerce businesses end up on a retained engagement of 10 to 15 hours a month within the first 3–6 months of starting with a fractional. The hourly entry-point is useful as a low-commitment trial period before either side commits to a retainer.

The honest framing of fractional bookkeeping is that software and people are complementary infrastructure, not substitutes. Software automates the matching layer that used to consume hours per cycle. The fractional bookkeeper handles the judgement, accountability, and credential layer that software cannot. A business that has automated reconciliation but not engaged a credentialled practitioner is leaving the higher-leverage work undone — the GST notice that needs a competent response, the marketplace dispute that needs to be filed inside the window, the monthly close that needs sign-off. ReconPe operates a curated directory at /find-a-bookkeeper/ for finance teams that want to engage a vetted CA or fractional bookkeeper directly. Verification is manual; engagement terms are between the client and the practitioner; ReconPe takes no commission and stays out of the money flow. The directory exists because the work that a fractional bookkeeper does — the judgement that follows the matching — is what makes the matching matter.