Soft Costs in Solar EPC: Stopping the Profit Leak

Quick Answer

Why This Matters for EPCs

Most EPC margin conversations start and end with hardware pricing, module cost, inverter cost, BOS pricing, because that's the number visible on every quote. But hardware margin has compressed across the industry to the point where it's rarely the actual lever left to pull, module and inverter pricing is largely set by the same supplier market every competing EPC buys from, leaving little room to differentiate on hardware cost alone. The real, recoverable margin sits in soft costs: the labour, delay, and rework that happen around a project rather than inside its bill of materials, and that never appear as a single tracked number on most EPCs' books. This guide breaks down five of the most common soft costs into real rupee figures for a mid-size, 40-project-a-month EPC, so the leak stops being abstract and becomes something to actually go fix.

1. Failed Site Visits: Rs 2.3 Lakh a Year

A failed site visit is one that doesn't convert to a sale, but still costs fuel, salesperson time, and often a pre-visit prep hour, sometimes a design pass that never gets used. For an EPC closing 40 projects a month, a realistic mix of 80 to 100 total site visits a month means 40 to 60 of them don't convert. At an average fuel-plus-time cost per visit, that adds up to roughly Rs 2.3 lakh a year, spent on visits that produced nothing.

This cost is largely a function of how much a salesperson can offer on the visit itself. A visit that ends with a real, customer-ready 3D proposal converts at a meaningfully higher rate than one that ends with "someone will follow up," which directly reduces how many visits fall into the failed category in the first place. Reslink's mobile design workflow lets a salesperson complete the 3D design, run the shadow simulation, and hand the customer a proposal before leaving the property, closing the gap between visiting and converting rather than just visiting faster.

2. Manual Proposal Preparation: Rs 1.6 Lakh a Year

Every manually built proposal, sized in one tool, priced in a spreadsheet, formatted in a document editor, takes real hours to assemble, and often gets rebuilt when a spec changes mid-conversation. Across a 40-project-a-month EPC, that adds up to roughly Rs 1.6 lakh a year in salesperson and back-office time spent formatting and re-formatting proposals rather than talking to customers. The full rupee breakdown of manual quoting versus proposal software is covered here, including the close-rate difference, not just the time cost.

The fix isn't faster manual work, it's generating the proposal from the design itself rather than as a separate document. Reslink builds the CAPEX and OPEX proposal directly from the completed 3D design, with yield simulation, rupee ROI, and subsidy calculation included automatically, and delivers it over WhatsApp before the visit ends.

3. BOM Rework: Rs 1.2 Lakh a Year

BOM rework happens when the procurement order doesn't match what actually gets installed, most often because the electrical bill, cable lengths, distribution boards, earthing conductors, was estimated by hand from a 2D layout instead of calculated from the real string-run geometry. Reordering, waiting on a corrected delivery, and the labour lost sorting it out on-site adds up to roughly Rs 1.2 lakh a year for a 40-project-a-month EPC.

The specific fix here is automated Bills of Electrical (BOE) generation, not a generic BOM tool. Reslink generates the BOE directly from the 3D model, so cable lengths and equipment counts reflect the actual design rather than a manual estimate, and it regenerates automatically if the design changes, rather than requiring someone to remember to update a spreadsheet. Bills of Structure, mounting rails, purlins, clamps, generates the same way from the same model.

4. DISCOM Delay Overhead: Rs 1.8 Lakh a Year

A Single Line Diagram in the wrong format, or missing a document a specific DISCOM requires, triggers a resubmission cycle that can add days or weeks to a project timeline. That delay isn't free: it's staff time spent chasing the resubmission, and often a customer getting anxious enough to call a competitor. Across a year, that overhead runs roughly Rs 1.8 lakh for a 40-project-a-month EPC, the largest single leak on this list.

Reslink generates the SLD, layout drawing, and string drawing in DISCOM-accepted format directly from the 3D design, no external consultant reformatting a US- or generic-format diagram before it can be submitted. The same design that produced the proposal produces the bank- and DISCOM-ready documentation.

5. Change Order Rework: Rs 1.5 Lakh a Year

A customer request, or a site condition discovered only once installation starts, triggers a change order, and in a manual workflow, that means separately updating the design, the BOM, and the proposal, three documents that don't talk to each other. How EPCs are handling this by revising design, BOM, and proposal on-site instead of days later is covered here. Left as a manual process, this rework runs roughly Rs 1.5 lakh a year across a 40-project-a-month EPC.

Because Reslink's BOM and proposal both generate from the same 3D model rather than as separate documents, a design change propagates automatically instead of requiring three manual updates. A revision made on-site during a customer conversation updates the BOM and the proposal in the same motion.

How This Scales With Project Volume

The Rs 8.4 lakh figure above is calculated for a 40-project-a-month EPC. Scaling it linearly by volume gives a rough sense of where a smaller or larger EPC would land, though in practice the scaling isn't perfectly straight-line: a smaller EPC often has less specialized staff to absorb DISCOM liaison work efficiently, while a larger EPC can sometimes dedicate a person to just that function, which changes the per-project cost in either direction.

Treat the 20 and 80-project figures as a starting reference point, not a precise projection. The real number for your specific EPC depends on your actual failed-visit rate, proposal turnaround time, and how manual each of the five processes currently is, which is exactly why the methodology below matters more than the top-line number itself.

How to Actually Measure Your Own Numbers

Two of the five categories are worth measuring precisely rather than estimating, since they're the most fixable and the easiest to get wrong by guessing.

Failed-visit rate. Pull two numbers from your CRM or lead log for a recent month: total site visits conducted, and total visits that resulted in a signed contract. Divide converted visits by total visits to get your real conversion rate, then subtract that from 100% to get your failed-visit rate. Most EPCs guess this number is lower than it actually is, since the visits that convert are more memorable than the ones that don't. If you don't currently track visits and conversions as two separate logged numbers, that gap is itself worth fixing before trying to calculate anything else on this list.

BOM error frequency. Go back through your last ten completed projects and count how many required a second material order, of any size, after the original procurement was placed. A second order for even one item counts. Divide that count by ten to get your BOM error rate. If it's higher than one in ten, the electrical bill is likely still being estimated rather than calculated from actual design geometry, which is the root cause covered in the BOM rework section above.

Both numbers take under an hour to pull together, and both are more reliable than a gut estimate, since memory tends to round in the EPC's favour on exactly the kind of failures nobody wants to dwell on.

What EPCs Should Do Now

Common Mistakes EPCs Make With Soft Costs

• Treating soft costs as unavoidable overhead rather than a fixable process problem with a real rupee number attached.
• Watching hardware margin closely while never totaling these five categories against actual project volume.
• Assuming one fix closes the whole leak. These categories compound, a slow manual proposal increases the odds a site visit fails to convert in the first place.
• Blaming an individual employee for rework that's actually a tooling or workflow gap, not a performance issue.
• Not re-checking this number as project volume grows. A soft-cost leak that was tolerable at 10 projects a month becomes a real, visible number at 40. The same scaling problem shows up across a wider set of execution mistakes covered here.

Where Reslink Fits Across All Five Leaks

Each of the five costs above traces back to the same root cause: design, proposal, BOM, and documentation living in separate tools with a manual handoff between each one. A failed site visit often means a salesperson leaves without anything to show the customer immediately. A manual proposal means redoing math and formatting that could come from the design itself. BOM rework means the electrical bill was estimated, not calculated from the actual design. DISCOM delay usually means an SLD in the wrong format needing reformatting. Change order rework means a mid-project revision has to be manually re-propagated across separate documents instead of one.

Reslink's mobile 3D design workflow addresses that shared root cause rather than treating each symptom separately. A salesperson completes the design, runs the shadow simulation, and generates a proposal on-site, closing the failed-visit and manual-proposal leaks together. The Bills of Electrical and Bills of Structure generate automatically from that same 3D model, closing the BOM-rework leak. The SLD generates in DISCOM-accepted format from the same design, closing the DISCOM-delay leak. And because every document traces back to one model, a mid-project change propagates automatically instead of requiring separate manual updates, closing the change-order leak.

None of this requires a separate fix, or a new hire, for each leak individually, it's one workflow change addressing all five at once.

See exactly how one 3D design connects to every downstream document → Book a demo

Frequently Asked Questions

Q1. How much do soft costs really cost a solar EPC in India?

For a 40-project-a-month EPC, five common categories, failed site visits, manual proposals, BOM rework, DISCOM delay overhead, and change order rework, add up to roughly Rs 8.4 lakh a year. The exact number scales with project volume and varies with how manual your current workflow is.

Q2. Which soft cost is usually the biggest?

Failed site visits and manual proposal preparation together typically account for close to half the total, since they compound, a slow, manual proposal after a site visit increases the odds that visit doesn't convert in the first place.

Q3. Are soft costs the same as overhead?

Not exactly. General overhead, rent, salaries, admin, is largely fixed regardless of process. Soft costs specifically are the labour, delay, and rework tied to how a project actually gets executed, which means they're addressable by changing the workflow, not just by cutting expenses.

Q4. Does fixing BOM rework mean buying a separate BOM tool?

Not necessarily. The actual fix is generating the Bills of Electrical (BOE) automatically from the 3D design itself, rather than estimating cable lengths and equipment counts by hand from a 2D layout, so the bill reflects the real design instead of an approximation.

Q5. How do I calculate my own soft-cost total?

Estimate your failed-visit rate, average proposal turnaround time, BOM error frequency, typical DISCOM resubmission count, and change-order frequency against your actual monthly project volume. The Rs 8.4 lakh figure here is illustrative for a 40-project EPC, a smaller or larger EPC's real number will differ proportionally.

You May Also Like

• Most Common Mistakes Solar EPCs Make — and How to Avoid Them
• Solar Proposal Software vs Manual Quotation: The India EPC Math
• Live Design Revisions