Read Repeated Percentage Changes Before an AI Growth Summary

A passage that describes repeated percentage change needs more than a percentage and an enthusiastic adjective. You need the starting amount, the base used at each step, the number and duration of the steps, and any adjustments between them. Before asking AI to summarize a growth claim in a business book, turn the supplied description into a small sequence. That sequence makes the calculation visible without treating it as a prediction.

This guide uses six original fictional cases with abstract marker units. They are not quotations, manuscript extracts, investment returns, population observations, client records or observed AI responses. Exact local fraction arithmetic checks the authored examples only. No model, business method, professional review or author endorsement was evaluated. The reading workflow helps distinguish what a supplied rule calculates from what a real source would need to establish about its application.

Identify the base before calculating the second change

A percentage increase of ten percent means an addition equal to one tenth of a specified base. When the next increase uses the new amount as its base, the second addition can differ from the first even if the percentage remains unchanged. If the source instead applies both additions to the original base, it defines a different calculation. Preserve that distinction before reducing a passage to a sentence about total growth.

OpenStax's Algebra 1 discussion of the growth factor explains the constant multiplier associated with equal input steps in an exponential relationship. See https://openstax.org/books/algebra-1/pages/5-4-3-exponential-change-the-growth-factor. This guide uses only that introductory multiplier and step vocabulary, read October 30, 2026. It adopts none of the source's worked examples, exercises or biological scenarios. The finite examples below were authored independently for this reading task.

In a note, name the base in words as well as symbols. Write ten percent of the previous step's ending amount, or ten percent of the original amount, according to what the supplied rule actually says. If the base is missing, mark it as unresolved. An assistant should not choose a convenient interpretation merely because a business chapter uses the word compound or a chart appears to rise quickly.

Case A: two ten-percent steps differ from two original-base additions

Fictional case A begins with 200 abstract marker units. Its first rule increases the current amount by ten percent at each of two steps. The sequence is 200, 220, 242. The two additions are twenty and twenty-two units. The total increase is forty-two units, or 21 percent of the original 200. Multiplying by 1.1 twice answers this explicitly stated rule.

A separate rule adds ten percent of the original 200 at each step. That sequence is 200, 220, 240, with two equal additions of twenty. Its total increase is forty units, or twenty percent of the original amount. The final values differ by two units. The difference comes from the base used for the second addition, not a dispute about how to add the same two quantities.

Ask an assistant to show each intermediate amount and identify the base for each percentage. A final answer alone can hide a changed rule. A useful explanation says that the current-base sequence ends at 242 while the fixed-original-base sequence ends at 240. It does not call the two-unit difference a realized benefit, a business loss or evidence that a particular operating strategy works.

Separate a total change from a per-step statement

A total percentage change compares an ending amount with a starting amount. A per-step percentage describes the multiplier applied at each defined step. They are related under a declared rule, but the total is not generally obtained by copying one rate into a sentence with a longer time span. The number of steps and the reference base remain part of the calculation.

Write a short question that says which quantity you want. Do you want the ending amount after the supplied steps, the total relative change from start to finish, or a constant per-step factor that reproduces two endpoints? Those are different tasks. The same inputs may answer more than one of them, but an explanation should label the result rather than let the word rate carry every meaning.

A source can also use step labels without claiming that they represent time. A step might mean a round in a hypothetical example, a batch index or a stated interval. Keep the actual definition. Replacing an abstract step with a year creates a temporal claim the source may never have made. None of this guide's marker sequences is a forecast for an organization or a recommendation about an actual financial decision.

Case B: equal percentage increases and decreases need not cancel

Fictional case B begins with 200 marker units, increases the current amount by 25 percent and then decreases the resulting amount by twenty percent. The first step produces 250. The second retains four fifths of 250, returning to 200. Here the factors 1.25 and 0.8 multiply to one. The paired percentages differ because the two changes use different bases.

A separate comparison increases 200 by 25 percent and then decreases 250 by 25 percent. Its ending amount is 187.5, a decline of 12.5 units from the original 200. The total relative change is negative one sixteenth, or a 6.25 percent decrease. Calling equal percentage numbers equal opposite amounts would misread the second base. Both sequences are fictional arithmetic, not an observed gain or loss.

To undo a nonzero multiplier, the required inverse factor is its reciprocal. That statement identifies an arithmetic relationship; it does not promise that a real process can be reversed or restored. A reader should distinguish undoing a numerical transformation from undoing a historical event, recovering lost information or returning a business to an earlier condition. Those applications require evidence beyond a pair of factors.

Do not average changing percentages before checking their product

When the current amount changes at each step, the ending amount depends on multiplying the supplied factors. An arithmetic average of the percentage numbers answers a different summary question. Applying that average repeatedly can produce a different endpoint from the original sequence. Whether an average is useful depends on the question and the declared convention, rather than on how familiar the average sounds.

An assistant can make this distinction explicit by returning two labeled calculations. The first uses the actual authored factors in order. The second applies an explicitly chosen summary factor to the same starting amount for the same number of steps. Compare the results without calling one a software bug or a deception. They calculate different rules; any diagnosis of a real explanation would require its actual text.

If a source supplies an equivalent constant factor, identify how it was defined. A factor that reproduces the endpoints does not recover all intermediate amounts or explain their causes. Many sequences can share an endpoint. An endpoint summary may be useful for a narrow comparison while leaving volatility, timing and mechanisms unresolved. This guide does not estimate uncertainty, select a real model or recommend a rate for a business plan.

Case C: the arithmetic mean of opposite changes can hide an endpoint difference

Fictional case C starts with 100 marker units and applies a fifty-percent increase followed by a fifty-percent decrease, each to the current amount. The sequence is 100, 150, 75. The arithmetic mean of the two percentage numbers is zero, but the endpoint is 25 units below the start. The product of factors is 1.5 times 0.5, or three quarters.

Applying a zero-percent step twice would instead leave the amount at 100. That is a different path and a different endpoint. A constant positive per-step factor that exactly matches the original two-step endpoint would have a square equal to three quarters. This guide leaves that relationship in exact squared form rather than introducing rounded roots as if they were exact supplied inputs.

The case also separates mathematical endpoint equivalence from an explanation of what happened. A statement that two paths end at the same value would not establish that they have the same intermediate values, causes or practical effects. Here even the simple arithmetic-average replacement fails to match the endpoint. Ask an assistant to retain the original sequence so the reader can see which information an average omits.

Preserve step duration when a percentage is attached to time

A multiplier per week is different from the same multiplier per month. To repeat a rule across a longer interval, you need a definition of the step and a compatible number of repetitions. Calendar labels, elapsed durations and business periods can introduce further questions. Do not substitute a number of days for a number of reporting periods without the source's actual convention.

A finite numerical example can state its own equal steps and avoid calendar ambiguity. For a real book passage, preserve the time unit, start and end basis, and whether a statement is observed, hypothetical or projected. If the duration is missing, the total change may still be calculable from supplied factors while a per-time-unit claim remains unresolved. Those are separate levels of what the input supports.

This distinction matters when a summary changes from over the stated period to every period. The first can describe one supplied comparison; the second asserts a repeatable rule. A multiplication that follows the stated example does not establish the rule will continue. Ask what source evidence supports extending it, and keep the extension labeled as an assumption when no such evidence is supplied.

Case D: one three-step factor differs from three repetitions of that factor

Fictional case D begins with eighty marker units. One authored rule multiplies the current amount by 1.25 for each of three equal abstract steps. Its amounts are eighty, one hundred, 125 and 156.25. The combined factor is 125 divided by sixty-four, or 1.953125. The final amount is therefore 95.3125 percent above the initial eighty under this finite rule.

Another authored statement applies one factor of 1.25 over the entire three-step span. Its endpoint is one hundred, an increase of twenty units or 25 percent. The endpoint difference between the two rules is 56.25 units. The same numeral 1.25 appears in both descriptions, but the number of applications differs. Neither result supplies a real period, an observed trend or evidence that the larger result is attainable.

Ask the assistant to echo the scope of the factor before calculating. A useful response states per step or over the whole span alongside the factor, then gives the number of applications. If a source leaves that scope unclear, retain both labeled interpretations rather than choosing the more dramatic result. The purpose of the comparison is to expose a reading question, not to manufacture a favorable outcome.

Locate additions and removals between multiplier steps

Repeated multiplication is only part of a rule if the supplied description also adds or removes an amount between steps. A fixed adjustment can make the timing of operations matter. Keeping the same factors and adjustment amount while moving the adjustment to a different point can produce a different endpoint. Preserve the sequence of verbs as well as the numerical values.

Write the operation order on separate lines: multiply, adjust, multiply again. If the adjustment's location is unknown, do not silently put it at the beginning or end. A summary that says after adjustments may be too vague to reproduce the example. Ask the source question directly: which base receives the next percentage, and does that base already contain the added amount?

When no adjustments occur, multiplication of scalar factors has familiar arithmetic properties, but those properties should not be generalized to every process containing percentage language. A real sequence may change membership, units or scope along the way. This guide does not inspect any actual operating system, pricing policy, account or model. It uses a small authored adjustment case to show why sequence deserves its own field in a reading note.

Case E: the same added amount has a different result when its position changes

Fictional case E begins with 100 marker units, multiplies by two, adds ten units and then multiplies by one half. The steps give 100, 200, 210 and 105. A second sequence begins with 100, multiplies by one half, adds ten and then multiplies by two. It gives 100, fifty, sixty and 120. Both use the same two factors and the same fixed addition, but their endpoints differ by fifteen units.

With no added ten, either pair of multipliers would return 100 because two times one half equals one. The fixed addition is scaled by the factor that follows it: five units of endpoint contribution in the first sequence and twenty in the second. That explains the difference using the supplied rules. It does not establish a causal result in a real process or recommend moving any actual adjustment.

Ask AI to preserve the placement of the addition in its explanation. If it reduces the example to net zero percent before considering the fixed amount, it has lost an operation. If it attributes the larger endpoint to a company's skill, it has added a business story. The checked example supports a narrow conclusion about the two authored sequences and leaves any real-world application to its own evidence.

Keep an unknown factor unresolved even when other steps are known

A sequence may be partly specified. Knowing the first multiplier, the starting amount and the number of positions does not necessarily determine the endpoint. An assistant should identify which factor is missing and retain a symbolic relationship when useful. Replacing an unknown factor with one assumes no change; replacing it with zero assumes elimination. Neither follows from missing information alone.

Labeled hypothetical completions can show why a missing factor matters. They should remain clearly separate from an estimate of the missing value. Choose simple completions to expose the dependency, not to create an optimistic and pessimistic forecast for an actual organization. A reader can then ask for the source's missing definition without mistaking the demonstration for an inferred answer.

An endpoint alone may also fail to identify the intermediate factors uniquely. A product can be constructed from many pairs. If the reading question concerns when change occurred or which step contributed it, the final number is insufficient. Keep the distinction between reproducing an endpoint and reconstructing a path. The latter needs information about the actual steps, their ordering and any adjustments.

Case F: a known first step leaves the complete endpoint unknown

Fictional case F begins with 120 marker units and a first multiplier of 1.1. The amount after that step is 132. A second factor is unknown, so the complete endpoint can be written 132 times the unknown factor. The starting amount and first factor do not determine a number for the full sequence. No missing factor was measured or inferred in this teaching case.

If the second factor were hypothetically 0.9, the endpoint would be 118.8. If it were hypothetically 1.2, the endpoint would be 158.4. The two hypothetical endpoints differ by 39.6 units. Neither completion is adopted. A response that substitutes one as the actual result would change the record; a response that says nothing at all is calculable would miss the known intermediate amount of 132.

The case also shows why counting known fields is not a comprehension score. One missing factor can determine the entire remaining calculation, while a complete numeric list can still have unclear meanings or units. Ask for an answer organized by question: the first-step amount is determined; the complete endpoint is unresolved; two labeled completions illustrate its dependence. That is more useful than one unqualified answer.

Ask AI for a sequence explanation with a bounded purpose

An original prompt for this task is: use only my supplied rule and inputs; state the starting amount, base of each percentage, factor, step unit and operation order; show intermediate amounts; distinguish total relative change from a per-step number; preserve unknown factors and label hypothetical completions; explain no actual business outcome, author method or forecast unless my inspected source establishes it. This is a proposed teaching prompt, not a tested guarantee of model behavior.

Review the response against the exact reading question. Did it apply each percentage to the specified base? Did it add percentage numbers where the rule required multiplication? Did it turn an equal decrease into a claimed reversal? Did it change one factor over a span into a factor repeated at every step? Did it move a fixed addition or fill an unknown with a neutral multiplier? These are concrete changes a reader can locate.

Then inspect the interpretation separately. Correct arithmetic does not show that the source's model is empirically supported, causally identified or transferable to another setting. It also does not show that an author endorsed the assistant's suggested application. Keep the source locator and the original rule with the note so a later passage, correction or definition can be incorporated without reconstructing the calculation from a polished summary.

Connect the repeated-change question to the right companion reading task

Use the equation-reading guide when the problem is grouping, symbol definitions or an undefined expression. Use the chart-reading guide when a scale, axis or date spacing creates the ambiguity. This guide's distinct task is the sequence of bases and factors across repeated changes. A chart may display that sequence, and an equation may encode it, but neither replaces the need to identify what receives each percentage.

Use the reported-precision guide when the source's decimal display and calculation precision are unclear. The exact fractions in this guide are an authored teaching choice rather than evidence that a real source has unlimited precision. If an actual passage gives rounded percentages, retain that fact. Do not invent extra digits to make a total match or diagnose an author's error without inspecting the stated rounding and calculation rules.

End a reading note with the missing question that could change the explanation. You may need the percentage base, the length of a step, the number of applications, the timing of an adjustment or the source of an empirical claim. If those are already clear, a short exact sequence may be sufficient. Further prose should explain the substantive evidence question rather than repeat the arithmetic as if repetition made a forecast more credible.

Frequently asked questions

Do two ten-percent increases equal a twenty-percent total increase?

It depends on the base. In case A, two increases applied to the current amount turn 200 into 242, a 21 percent total increase. Two additions each based on the original 200 instead end at 240. State the rule before comparing the totals; neither case is an observed business result.

Does the same percentage increase and decrease return to the start?

Not generally when each uses the current amount. Case B's increase and decrease of 25 percent end below the starting amount. A 25 percent increase followed by a twenty-percent decrease returns to the original under the authored rule because the factors are reciprocals. That arithmetic is not a promise that a real process can be reversed.

Can I apply the average percentage repeatedly instead of the original changes?

Treat that as a different calculation unless the chosen summary is defined to answer the same question. Case C's fifty-percent increase and decrease have arithmetic mean zero but end at 75 from a start of 100. Repeating zero-percent steps leaves 100. Preserve the factors and endpoint comparison rather than substituting an unlabeled average.

Why does per-step wording matter if the factor is unchanged?

It determines how often the factor is applied. Case D repeats 1.25 three times in one rule and applies it once across the whole span in another. The resulting endpoints differ. A number without its interval or step scope is insufficient for a determined repeated-change explanation or an actual temporal claim.

Can a missing factor be treated as one?

Only as an explicitly labeled added assumption. Case F retains its second factor as unknown while calculating the first-step amount of 132. Its two hypothetical completions produce different endpoints and neither recovers the original missing input. A convenient complete result should not conceal a gap in the supplied rule.

Are these examples taken from Connor Robertson's books or an actual model response?

No. The sequences, abstract units and teaching prompts are original fiction. No book passage, manuscript, chapter method, author quotation, observed model output, client result, professional review or endorsement was inspected or attributed. Local exact arithmetic checks the supplied cases and provides no empirical forecast, investment guidance or operating recommendation.