8 Questions Before Showing Projected Solar Savings

Projected savings often arrives in a solar proposal as the most polished number and the least inspectable one. A customer sees a large lifetime total or a first-year reduction, while the supporting chain sits across bills, tariff records, design versions, production models, financing documents, and analyst assumptions.

That chain is the real product of the analysis. If one material link belongs to another meter, an old design, a cash price the customer is not paying, or an export rule that does not apply, the displayed total can be internally consistent and still answer the wrong question.

The following eight questions give solar sales and design teams a release test. They do not produce one correct savings number for every project. They tell the team whether the stated scenario has a traceable basis and whether the customer can understand what may change it.

1. Does the load record belong to this customer decision?

The load record should identify the meter, premises, customer, billing periods, gaps, tariff, and expected operating changes relevant to the proposal. A clean annual total is insufficient when it combines accounts, omits a material season, or hides a future load change. Match the evidence depth to the decision the projection will support. It should also record who confirmed the identity.

Confirm that each bill or interval file belongs to the premises in the proposal. Record missing months and estimated readings where visible. A summary without the underlying periods can hide a mixed set of accounts or a partial year.

Load shape can matter as much as annual consumption. A business that operates mainly during solar hours may use generation differently from a building with evening demand. Time-dependent rates, export compensation, demand charges, storage operation, and seasonal closure can all make a single annual number insufficient for a detailed scenario.

Ask what the projection is meant to support. A first screening conversation may use limited customer information with an explicit low-confidence label. A finance or contract decision needs the evidence appropriate to that consequence. Do not quietly reuse a screening load after the customer supplies better data.

Future changes remain assumptions until supported. Planned electric vehicles, heat pumps, new machinery, tenant changes, and facility expansion may belong in a separate scenario. Keep historic consumption and customer-planned use distinct so a later reviewer can see which part came from a document and which part came from a conversation.

The customer should be able to recognize the basis in one line: “Model uses the identified meter and the billing periods listed.” If that sentence cannot be written, resolve the account before displaying projected savings.

2. Is the tariff the one that will value the energy?

The savings model needs the tariff that applies to the identified meter and scenario, including material energy periods, tiers, demand components, fixed charges, riders, taxes, export treatment, and eligibility conditions. The method should distinguish charges that may change with grid purchases from costs that remain under the stated project scenario. Keep the source, effective date, and review owner in view.

The tariff is not merely the price printed in large type on a bill. A bill shows charges for its period; it may not describe every future rule or condition.

Use a current utility or regulator source for the applicable jurisdiction and record the access date. A bill can show what the customer was charged during a period, but it may not describe future rates or every condition. When the tariff is unclear, qualify the model or route it for specialist review.

Do not assume every generated unit offsets the highest retail rate. Some energy may be consumed on site, some exported, some curtailed, and some shifted through storage under modeled controls. The applicable rules determine the value assigned in the financial scenario. A production total alone does not establish that value.

Separate the current tariff from future escalation. The current source is an observed input. Escalation is a model choice unless a controlling schedule establishes a future change. Show a neutral sensitivity view rather than burying a favorable escalation inside one lifetime total.

The U.S. Department of Energy’s consumer solar resources direct customers toward current information and consumer-protection materials. For projects outside the United States, use the equivalent primary utility, regulator, and contract sources. Jurisdiction is part of the claim.

3. Does the production model match the design being sold?

The production model must identify the design revision being offered, including material module count, orientation, equipment, shading, and configuration assumptions. Link the proposal to a production-run identifier and define rerun triggers. A newly dated financial page does not make an older energy result consistent with a design that changed. Any material design revision should trigger another production run before release.

Savings scenarios often remain attached to earlier layouts. Review both files together before release instead of trusting the latest timestamp on the proposal.

Check the physical chain: roof or site information, array placement, selected module and inverter assumptions, orientation, shading treatment, losses, availability or curtailment assumptions where relevant, and weather or solar resource data. The level of detail should fit the decision, but the model should never imply site verification that did not occur.

The National Laboratory of the Rockies (NLR, formerly the National Renewable Energy Laboratory/NREL) provides the PVWatts calculator, which produces an estimate from defined system and resource inputs. Its V8 API documentation exposes those input fields. These official resources support a simple discipline: an annual production number has a model basis, and changing that basis can change the output.

Ask the designer to approve the production input set before it becomes customer-facing. Approval should mean fit for the named proposal use, not a guarantee of future generation. Record unresolved site conditions and define which new evidence triggers a rerun.

If the proposal presents multiple system options, keep each option paired with its own production result. Reusing one output across different placements or equipment packages may create a tidy table but breaks the evidence chain.

4. How does generated energy become financial value?

The valuation method should state how modeled generation interacts with on-site consumption, exports, time periods, demand, fixed charges, storage, and other material tariff components. It must show which avoided or credited charges create the projected value. A customer should understand the mechanism without needing the analyst’s private spreadsheet. The explanation should identify the period and customer behavior the model assumes.

Write that method in ordinary language. One unit of production may carry a different modeled value from another because timing, use, and export treatment differ.

For a simple netting structure, the explanation may be short. For time-varying tariffs, commercial demand, export caps, or batteries, the interaction can require interval data and dispatch assumptions. Show what the model actually represents. Do not use a sophisticated label to hide a rough approximation.

Keep energy savings separate from other claimed value. Resilience, property value, environmental attributes, tax effects, and operational benefits require their own definitions and evidence. Adding them to a single “savings” total without visible boundaries prevents fair comparison.

Use a reconciliation check. Starting from the customer’s historic bill, identify which modeled charges can change, which remain, and which new project costs appear. The resulting view may not match a simplistic consumption-times-rate calculation, and that is useful. It shows the customer where the money comes from.

Define bill savings as the difference between paired modeled bills without and with the project under the same stated load and tariff basis, except for explicitly declared project changes. If export credits already reduce the with-project bill, do not add them again as a separate benefit. Keep PV energy at its declared AC or DC boundary distinct from utility imports and exports; storage losses and grid charging must follow the modeled dispatch.

Show net owner cash flow separately from bill savings, with debt, operating expenses and any conditional incentive included once under a declared ledger. First-year nominal savings, cumulative nominal savings and discounted present value answer different questions; label the period, currency basis and discount assumptions rather than calling all three lifetime savings.

A solar savings calculator for homeowners can help explain inputs at a high level. A proposal release still needs project-specific sources, dates, and a calculation method suited to the actual tariff and design.

5. Are financing and incentives represented as current, conditional inputs?

Financing and incentive inputs should come from current controlling documents, carry eligibility and expiry conditions, and remain separate from physical energy performance. Record cash price, financed amount, fees, interest terms, payments, and incentive treatment. Do not place financed payments beside cash-based savings as though both views use the same cost structure. Every material term needs a named source and date.

Financing changes customer cash flow. A payment figure does not by itself describe total cost, and an incentive amount does not establish that this customer qualifies.

Incentives, tax treatment, and credits are jurisdiction-sensitive and time-sensitive. Confirm effective dates, eligibility, ownership requirements, caps, assignment rules, and interaction with financing using current primary authority. The salesperson should not decide that a customer qualifies for a tax benefit.

Show gross project economics and the selected financing view separately when that improves clarity. A customer may then compare the physical project and the proposed payment arrangement without confusing an energy outcome with a credit product.

The Federal Trade Commission’s consumer solar guidance asks buyers to inspect claims, contracts, and financing. The DOE homeowner solar guide also directs buyers to evaluate bids and financing. These sources reinforce disclosure and comparison, not any particular projected saving.

Set expiry dates for financing and incentive claims. If the proposal is reopened after the document expires or policy changes, hold the financial page until the controlling source is refreshed.

6. Which operating costs and exclusions sit outside the headline?

The model should list every material operating cost as sourced, assumed, or excluded. Relevant items may include monitoring, maintenance, insurance effects, financing cost, replacements, degradation or downtime effects, taxes, or subscriptions. Degradation and downtime generally affect energy output rather than constituting a second cash expense; avoid counting an output reduction again as an added charge. An exclusion does not mean the cost is zero; it defines the boundary of the projected savings figure. The customer must see those statuses beside the figure, in readable language there.

Savings is a net concept only when the calculation states what it nets. Two offers can display different totals simply because their included costs differ.

Use three statuses: included with source, included as a labeled assumption, or excluded from the calculation. The distinction lets a customer compare two offers whose headline totals use different boundaries. It also stops a team from adding every optimistic benefit while omitting plausible costs.

Equipment warranties and service plans are not the same as a cash reserve. Read the actual terms before assigning value or claiming coverage. A warranty may have exclusions, procedures, labor treatment, transfer rules, or conditions. If the financial model includes a replacement or maintenance allowance, name its basis.

For commercial projects, operating assumptions may interact with asset ownership and contracts. Keep the solar proposal within its competence and give the customer’s finance, legal, tax, or insurance advisers a clear input register. An open issue is safer than a confident but unsupported zero.

The model owner should sign off on exclusions before sales sees the headline. A salesperson cannot explain a boundary that the analysis never recorded.

7. Does the scenario show uncertainty without manufacturing a range?

Show uncertainty through a few material, traceable sensitivities rather than an arbitrary high and low range. Each changed input needs a source, an explicit customer-defined label, or an honest assumption status. Keep the method and cost boundary consistent, and explain which evidence would help the customer choose between scenarios. Each scenario should end with the evidence needed for the choice.

An arbitrary range adds decoration rather than information. Choose sensitivity inputs because they can affect this project, not because the tool makes them easy to vary.

Useful candidates include consumption timing, future facility load, energy escalation, export value, production, operating cost, and financing structure. Change one or a small group transparently. Keep the calculation method and cost boundary consistent across columns.

Do not call the middle case “most likely” unless evidence supports a probability judgment. Planning case, current-input case, and customer-defined case are safer labels. Explain that the scenarios test decisions; they do not bound every possible future result.

Avoid false precision. If the underlying data is preliminary, a lifetime total shown to the nearest unit can make the model look measured. Round in a manner appropriate to the decision, state the model date, and let the evidence quality control the confidence label.

The scenario table should end with the variable to verify next. A range earns its space when it helps the customer decide whether to collect interval data, confirm a tariff, change the design, or compare financing. Otherwise the range is merely a wider sales claim.

8. Can another reviewer reproduce the customer-facing result?

Another qualified reviewer should be able to reproduce the displayed scenario from retained records without interviewing the original analyst. The record must identify the design, production run, customer load, tariff, financial terms, incentives, operating assumptions, calculation method, and proposal version. Unreproducible savings should not reach the customer. The retained record should also show its source dates and reviewer by name.

Reproducibility is the final release test. Preserve the working inputs and formulas, not only a screenshot of the headline.

Record dates, owners, and the software or workbook version where relevant. If a live data source can change, retain enough context to explain the value used at release.

Reproduction means running the stated method with retained inputs and a defined ownership and cash-flow boundary, not forcing every project into one calculation.

Perform an independent claim scan on the customer page. Find every currency figure, percentage, year, tariff statement, incentive, performance number, and comparative claim. Bind each to a current source or validated calculation. Remove or qualify anything that cannot survive that check.

Finally, compare the spoken sales script with the document. If the proposal says “modeled estimate” while the salesperson says “guaranteed savings,” the artifact has not controlled the claim. Training and release review belong to the same evidence system.

Build a customer-facing savings review packet

The eight answers need to travel together. A customer-facing savings review packet should identify the customer’s decision, project and calculation revision, electricity and tariff sources, system and production basis, ownership scenario, assumptions, exclusions, sensitive inputs, reviewer, permitted use, and next verification step.

The packet is smaller than the full analyst model. It does not hide the model; it gives the customer a readable map back to it. A presenter should be able to open the source or reviewer record for any material figure without turning the meeting into an audit of every cell.

Do not combine scenarios to make one clean headline. A future-load case, alternate tariff, different system size, or changed ownership structure should have its own identifier. The customer can compare options only when the design, energy, cost, and financial assumptions are internally consistent.

Use an evidence-label key on the page

Give every material input one of four labels:

A customer-supplied label does not establish independent verification. It means the customer supplied or confirmed it for the stated purpose. A future load remains a customer plan, not observed consumption. Project evidence also has a release boundary; preliminary site information is not final field verification.

Walk through one discrepancy without defending the number

Illustrative example, not a customer result: A customer asks why the projected savings changed after a revised roof layout. The earlier scenario used a larger preliminary array area. New site evidence removed part of that area, and the project team created a new layout and production model.

The presenter starts with the revision note. They show the old and new scenario identifiers, the site evidence that changed the usable area, the production output affected, and the current savings calculation. They do not imply that the first result was guaranteed or that the second is final beyond its evidence.

If the tariff or ownership assumption did not change, say so. If the revised system changes cost or finance inputs, route those fields through the appropriate reviewer and show them under the new scenario. The customer sees which part of the result moved and why rather than receiving a replacement headline with no history.

The next action might be to confirm an electricity record, choose between system options, obtain a reviewed ownership scenario, or wait for another project input. The presenter should name the evidence choice instead of pressing the customer to accept the latest estimate.

Give the presenter a stop script

When a load-bearing input cannot be verified, the presenter needs language that protects the conversation:

This result depends on an input we have not yet confirmed for your project. I can show where it enters the scenario and what it affects, but I should not present the current figure as decision-ready. The next useful step is to obtain or review the missing evidence, then issue a new labelled calculation.

That answer is more credible than inventing a range or quoting a market average without support. It also gives the customer a concrete next step. The actual route depends on whether the missing item concerns electricity, tariff, site, system, incentive, tax, financing, contract, or another specialist question.

Use the plain-language solar payback meeting guide when the customer needs the crossing-point concept rather than a full savings walkthrough. Keep payback, cash flow, tax, financing, and other metrics distinct.

Run the meeting in an evidence-first sequence

The sequence keeps the customer from anchoring on a savings headline before understanding the project and scenario behind it.

1. Confirm the customer decision and the option being discussed.
2. Show the current design and scenario identifier.
3. Identify customer records and statements, project evidence, assumptions and unknowns.
4. Explain how the production model connects to generated energy.
5. Show how the reviewed tariff and ownership scenario convert energy into modeled value.
6. Name exclusions, sensitive inputs, and questions owned by specialist reviewers.
7. End with the customer’s next evidence or option decision.

Do not start with the “best” scenario and explain limitations later. A detached headline may convey stronger certainty than the source and assumptions support. Put the basis on screen before the outcome and keep it visible while questions are answered.

When another quote is involved, use the nonequivalent solar quote comparison guide. A different savings result can come from system size, production, load, tariff, costs, ownership, incentives, exclusions, or calculation method. Compare the inputs before deciding which output is more credible.

Record the questions the customer could not answer in the meeting. Assign each to the customer, project team, utility or authority source, finance or contract reviewer, tax professional, or another appropriate owner. Do not let an open question become an unchanged assumption in the next proposal revision.

After the meeting, send the scenario identifier, evidence requested, decisions made, and next update. Avoid sending a detached screenshot of the savings figure. The follow-up should let another stakeholder understand the current basis without hearing the original explanation.

If the customer asks for a simpler page, simplify the presentation rather than removing the source labels. A short summary can still show the decision, scenario, three or four load-bearing inputs, limitation, and next step. Clarity and traceability are compatible.

Assemble the savings release record

Use a compact record for every customer-facing scenario:

This record should sit behind the proposal, not become a dense customer worksheet. Surface the items that change the decision and offer the supporting detail. A buyer gets a readable explanation; the company retains a traceable calculation.

Run two reviews before the customer meeting

The first review checks the model. Give the reviewer the load files, tariff record, design revision, production run, cost and financing documents, incentive authorities, assumption register, and customer-facing result. Ask the reviewer to reproduce the displayed value and identify every material number that lacks a retained source or calculation binding.

Do not let the analyst guide the reviewer through hidden steps. Questions are useful after the independent attempt, but the initial reproduction should test whether the record stands on its own. A step that lives only in someone’s memory will become a problem when the proposal is revised, the analyst is unavailable, or a customer asks for an explanation months later.

The second review checks the conversation. A sales manager or colleague should read only the customer page and explain what the projection includes, which facts came from the customer, which values are modeled, what can change, and which next document would improve confidence. The test catches language that is technically qualified in a footnote but absolute in a headline.

Reviewers should also inspect presentation order. A large lifetime figure shown before the method can anchor the conversation around an unsupported sense of certainty. Put the scenario label and time basis beside the number. Show first-year, cumulative, nominal, and discounted views only when their definitions are explicit and the customer decision needs them.

Record the review result with named exceptions. “Approved” carries little information. “Load and current tariff confirmed; export treatment remains provisional pending utility response” gives sales a usable boundary. It also tells the model owner what event requires the page to be updated.

Adapt the questions to the project type

A residential customer may have one meter and a simple operating pattern, yet financing and future consumption can dominate the discussion. Ask whether planned electrification belongs in a separate case. Keep lender terms visible. Explain which fixed bill components remain even when grid purchases fall.

A commercial building may need interval load, demand treatment, operating schedules, tenant responsibilities, and facility-change assumptions. The annual bill total might support a rough screen while being inadequate for a detailed demand or storage scenario. Let the business’s actual decision determine the evidence request.

A multi-site portfolio needs identity control before aggregation. Each meter, tariff, building, design, and production run should map to the correct site. A portfolio total can hide one location with bad data or a different rule. Release site-level records first, then show how they combine.

Storage adds another modeled layer. Battery operating logic, usable capacity, efficiency assumptions, reserve settings, equipment constraints, and tariff interaction can change the time and value of energy flows. Do not attach a battery uplift to a PV-only savings model or describe backup value as an energy-bill guarantee.

For leased premises, shared roofs, or third-party ownership, ask who pays each charge and who receives each benefit. A technically valid production scenario may not answer the contractual allocation question. Leave legal and tax interpretations to qualified advisers while giving them a clean technical and cash-flow record.

Control revisions after the meeting

Customer feedback often changes the scenario. They may correct a usage assumption, request a different financing offer, add storage, prefer another layout, or disclose a planned load. Treat each change as a model event rather than editing the proposal headline by hand.

Create a change note that identifies the request, source, affected inputs, new design or model run, reviewer, and superseded proposal. If the customer asks for a favorable hypothetical, label it as customer-defined and keep the current-evidence case beside it. A requested scenario is not a forecast.

Archive the released view. The team needs to know what the customer saw, not merely what the model says after later updates. When a result changes, explain the reason in terms of changed evidence or method. Quietly replacing a figure encourages confusion and makes later dispute review harder.

Set refresh triggers by claim. Tariffs, financing terms, incentives, equipment prices, and policy treatment can expire sooner than geometry or a method definition. A proposal reopened after its volatile inputs expire should return to review, even if its file date still looks recent.

Finally, close the loop with design and operations. If new site evidence changes placement or production, the proposal owner should receive the event. If a commercial assumption changes system selection, the designer should see it. Savings is downstream of both teams, so revision control cannot end at the sales document.

SurgePV supports 3D roof modeling, solar array layout, shading analysis, energy-yield and financial modeling, electrical workflow support, bill-of-materials output, and proposal generation. Results depend on source data, assumptions, equipment models, configuration, and review. Outputs support design and documentation workflows but do not replace approval by the responsible engineer, authority, lender, insurer, or utility.

Frequently Asked Questions

Can a solar company guarantee projected savings?

No responsible projection should be presented as a guaranteed bill reduction. Future savings can depend on weather, production, customer consumption, tariff changes, export treatment, equipment behavior, financing, operating costs, and policy. A company can document the model, sources, assumptions, and review, then update the scenario when material inputs change.

How many electricity bills are needed for a savings projection?

There is no universal document count that makes every projection sufficient. The analyst needs enough current information to represent the relevant load and tariff for the decision. Seasonal operations, interval-sensitive charges, tenant changes, electrification plans, and incomplete billing periods can require more detail than a simple annual total.

Should projected savings include electricity price increases?

Only when the escalation method is clearly labeled, supported, and appropriate to the jurisdiction and decision. Show the current tariff source and let the customer see how a different escalation assumption changes the scenario. Do not choose an aggressive future rate merely to create a more attractive lifetime savings figure.

What is the difference between production and savings?

Production is modeled electrical energy from the proposed system under stated design and performance inputs. Savings assigns financial value to how that energy interacts with customer use, tariffs, exports, financing, and other costs. The same production estimate can produce different savings scenarios when the customer’s load or electricity treatment changes.

When should projected savings be recalculated?

Recalculate when a material input changes, including the customer load, tariff, export rule, design, equipment, shading treatment, production method, project price, financing, incentive assumption, operating cost, or analysis period. The proposal should identify its input dates and revision so a superseded result is not reused in a later decision.