Pro versus Con Analysis

Goal

Pro versus Con Analysis answers a single decision, most often a binary one: adopt or not, sign or not, launch or not. The arguments already exist, but scattered, in the heads of the people concerned, and the risk is to decide on whichever ones happen to be present to the mind at the moment of choice rather than on the whole set. Benjamin Franklin, who left the earliest known description of the method in 1772, framed exactly that difficulty: as long as a decision is held "under consideration", all the reasons for and against are not present to the mind at the same time. The technique brings them all "together in one view" so that judgement rests on the totality.

The decision it supports concerns a single option, weighed on its own merits: it informs a go/no-go. The deliverable is the list itself, two columns of arguments for and against, optionally with a weight per argument and a resolved balance, reached either by summing the weights of each column or by Franklin's striking-out, which cancels two opposing arguments of comparable weight. This artefact is the record of why the decision was made: it can be shared, audited and revisited if the decision has to be reopened. The technique belongs to the decision analysis family, of which it is the lightest member.

Usage

When to use it

• A single go/no-go decision: adopt a tool or not, sign a contract or not, launch an initiative or not.
• Modest stakes, a quick decision: a formal method would cost more than the decision is worth.
• A need for transparency in a group: making the reasoning visible and open to challenge in a few minutes.
• Structuring a personal or small-group deliberation: laying the reasons out flat before deciding, Franklin's original use.

When not to use it

• Several options to compare: one pro/con list per option does not set them against each other, move to the decision matrix.
• Several weighted criteria and high stakes: a rigorous multi-criteria trade-off, move to multi-criteria decision analysis (MCDA).
• A decision about a change and its forces: analyse the driving and restraining forces rather than arguments, move to force field analysis.

Description

The procedure

The procedure Franklin described in 1772 under the name "moral or prudential algebra" remains, in essence, the method used today. BABOK lists it among the tools of decision analysis without saying more; the technique's own literature sets it out in six steps.

1. State the decision as a single proposition. One option, framed as a closed question ("adopt X?"). If the statement does not fit in one proposition, there is more than one subject and this is the wrong technique.
2. Draw two columns, Pro and Con. Franklin did this by dividing half a sheet of paper with a single line.
3. Generate the arguments. Franklin gave it "three or four days", adding reasons as they occurred to him. The decisive discipline here is honesty of generation: ideally a second person, or the advocate of the option compelled to argue the side they disfavour, writes the con column, because a list drawn up by someone who already leans towards the option is stacked from the outset.
4. Weight each argument, optional but strongly recommended, for example from 1 to 5 according to its importance in the light of the decision-maker's values. This is where stakeholder values legitimately enter the technique, and it is what forbids simply counting the columns: a single heavy con can outweigh six light pros.
5. Resolve the balance. Two equivalent ways. Franklin's striking-out cancels one pro against one con of roughly equal weight, "and where I find two, one on each side, that seem equal, I strike them both out", until one column prevails. Or you sum the weights per column and compare the totals. Franklin called the whole thing "moral algebra": the weights do not have the precision of real algebra, but setting them down and comparing them "separately and comparatively" is what lets one judge better and be less liable to a rash step.
6. Decide. The artefact informs the judgement, it does not automate it. The decision-maker keeps the call and records it.

The pitfalls

One-sided framing

With a single-option list, it is the author who decides what goes into each column and the weight of each argument. Confirmation bias, and more precisely myside bias, is well documented: people spontaneously generate far more arguments for the position they prefer than against it, and produce counter-arguments only when explicitly asked to (Nickerson, 1998). A pro/con list drawn up by a person who already leans towards the option is therefore stacked by default, and the insidious part is that it looks balanced because it has two columns. The remedy is procedural: give the con column to someone other than the advocate or compel the advocate to argue the disfavoured side explicitly.

Manufactured mirror arguments

The two columns invite a false symmetry: entering "cheap" among the pros, then inventing "expensive" among the cons to fill the other side. Each argument stands on its own weight and sits on one side only, where it bears. An attribute, price, speed, risk, is assigned to a single side, the one where it is a reason. If a moderate cost is a genuine reason for the option, it belongs among the pros, and it calls for no answering "expensive" con unless the expense is, separately, a reason against. A cheap / neutral / expensive scale takes on meaning only once there are three or more subjects ranked against one another, and at that point you have left pro versus con analysis for a decision matrix or a multi-criteria decision analysis. Manufactured balance is the commonest way this technique comes to lie.

Counting the columns

A list without weights tempts one to tally the columns. Franklin had already settled the matter through weighting and striking-out, which are the method itself. One argument can outweigh five. A list left unweighted must state that the balance is read from the importance of the arguments.

Retrospective rationalisation

Once a decision has been made on instinct, the sheet can serve to manufacture a rational-looking justification for it: the columns dress an emotional choice in the costume of analysis. Honesty of generation is the safeguard.

Poverty of imagination

The list is only as good as the arguments one managed to produce: a whole category of consideration that no one thought of stays invisible on the sheet, because there is no empty cell to flag it. That is the price of requiring no data, and a model helps most at this point.

AI considerations

The most useful application attacks myside bias where humans are weakest. You ask the model to play devil's advocate: to produce the strongest arguments against the side the author favours or against the option altogether. People under-produce counter-arguments until they are prompted, and the model is a tireless prompt. In the same vein, it surfaces candidate arguments the author would not have thought of, which attacks poverty of imagination, and it extracts arguments from supplied documents, a contract, a vendor sheet, an internal policy.

Three limits are firm. The model does not assign the weights: the weighting encodes the values of the decision-maker and the stakeholders, it is the human core of the technique and it cannot be delegated. Nor does it decide, and above all it must not manufacture mirror arguments: told to "balance the two columns", a model will dutifully invent symmetric cons to fill the shorter side, which is the defect to avoid. You ask it for arguments that bear. Finally, data sensitivity: a real decision often rests on confidential figures, a price, the terms of an agreement, a personnel matter, which are not fed into a model whose data handling is not controlled.

Examples

Adopt an invoicing tool or not

A fiduciary firm of about a dozen people, in the canton of Fribourg, is considering whether to adopt Bexio for its bookkeeping and invoicing. One option, a go/no-go. The six arguments retained are weighted from 1 to 5 by their importance to the firm, and the balance is read from the sum of the weights in each column.

Each attribute appears on one side only. The recurring cost is a genuine con and shows up once, on the right, with no "savings" pro placed opposite it to feign balance. Remote access is a genuine pro, with no "loss of control" con to mirror it. The columns are uneven in their wording: the balance comes from the weights. A cheap / neutral / expensive rating would make sense only if the firm were comparing Bexio against two other packages, which would be a decision matrix, a different technique.

Visualisations

The drawing sets the two columns side by side, each argument carrying its weight chip, and a balance strip sums each side and states the outcome. It carries the technique's central lesson: an argument sits on one side only, and only a column's total is telling. The two columns need not line up row by row, and nothing in it suggests that a pro answers a con face to face.

The colour choice follows the same restraint. Both sides share the same blue, structural frame and are told apart by their heading and their sign, not by a green-red coding that would sort the good and bad camps in advance. The conclusion alone takes the accent, because that is what the reader carries away.

Beyond two or three arguments a side, or as soon as a third subject enters the comparison, the two-column form stops serving. The question becomes a multi-criteria ranking, and the right representation is a grid, one row per option and one column per criterion. Moving from one to the other is the signal that the technique has changed.

Cost

The low cost across all three phases is the technique's reason for being and the criterion that selects it. A pro versus con that turns expensive has usually become a decision matrix.

Tooling

Paper or a whiteboard are enough in most cases: this is literally Franklin's tool, half a sheet and a line. A two-column table in a word processor does the same job.

A spreadsheet serves the weighted variant: it sums the weights per column and makes the weighting explicit, which cuts short the reflex to count the arguments.

A shared collaborative document or whiteboard suits a distributed team: the artefact is text, and it is shared without a wall or sticky notes.

Needing specialised decision-support software is a warning sign: you have usually outgrown pro versus con, and a decision matrix or a multi-criteria decision analysis is called for.

Sources

• IIBA, A Guide to the Business Analysis Body of Knowledge (BABOK Guide) v3, §10.16 Decision Analysis: the descriptive frame, which lists pro versus con considerations among the tools of decision analysis.
• Benjamin Franklin to Joseph Priestley, 19 September 1772 (Founders Online, National Archives): the primary source and the origin of the method, the two columns, the striking-out of equal reasons and "moral or prudential algebra".
• Irving L. Janis and Leon Mann, Decision Making: A Psychological Analysis of Conflict, Choice, and Commitment, Free Press, 1977: the "decisional balance sheet", the psychological formalisation of the balance idea, of which Franklin is the precursor.
• Raymond S. Nickerson, "Confirmation Bias: A Ubiquitous Phenomenon in Many Guises", Review of General Psychology, 2(2), 1998: the reference review that anchors the myside-bias pitfall.