Bullseye

Goal

The bullseye sorts a set of items by their proximity to a single goal. It handles a list of heterogeneous candidates, the initiatives of a programme, the features of a product or the ideas coming out of a workshop, which everyone considers worthwhile and which no common scale separates. The technique asks one question, the contribution to a stated goal. A group that finds everything important can still answer that one.

Prioritisation gathers several approaches with distinct mechanisms. Grouping files items into bands named and defined in advance. Ranking imposes a total order, from first to last. Time boxing and budgeting start from the container, a duration or an envelope, and let the resource decide. Negotiation brings together stakeholders whose interests diverge. The bullseye and dot voting are the family's two instruments of collective judgement: each collects a room's opinion on a question set in advance. The decision comes out of what each person places, a vote for dot voting and a distance to the goal for the bullseye, where negotiation draws it from an agreement between stakeholders.

The deliverable is the filled target, which is photographed or exported: the goal readable in the centre, each item on a ring and the placement fixed at the end of the session.

Usage

When to use it

• A single verifiable goal: one sentence whose achievement can be stated and against which every item is measured.
• Heterogeneous candidates: comparing initiatives with no common scale of cost or size.
• Scoping or design workshop: obtaining a prioritisation in the session, with no prior estimate.
• Mixed business and technical group: everyone answers the proximity question without knowing the build cost.
• Inflation of high priorities: the primary ring's capacity forces the trade-off a written definition fails to obtain.

When not to use it

• An execution sequence to establish: the target leaves open which of the primary items gets built first, so take ranking.
• Openly conflicting interests: the placement would record a balance of power, so run the negotiation before any item is placed.

Description

A named centre and three rings

At the middle of the target, a disc carries the goal, written out in full and visible throughout the session. Its wording decides the rest: a verifiable goal, "halve counter visits by the end of 2027", separates the items, where "improve the customer experience" draws them all towards the primary ring. One target carries one goal; where several goals compete, run one target per goal.

The three rings surround this disc and LUMA names them Primary, Secondary and Tertiary. The primary ring carries what serves the goal directly: remove the item and the goal is missed. The secondary ring carries what contributes through an intermediary, a prerequisite, a capability that makes the primary items possible or a partial improvement. The tertiary ring carries what relates to the goal from a distance. It says "far from this particular goal", which leaves the item valid with respect to another goal and a candidate for another target. This is the reading the group loses fastest, and the facilitator recalls it before the first placement.

The primary ring's capacity is the mechanism of the technique. The ring is drawn narrow, so the number of labels it accepts is bounded. A capacity of three to five items makes the trade-off real without making it arbitrary. Under that constraint, the group argues over what deserves one of the places instead of noting afterwards that fifteen items out of twenty are critical.

The target has one dimension, the distance to the centre. An item receives a single judgement there, where a matrix of importance against difficulty asks for two and a pairwise comparison asks for as many as there are pairs. That is what makes the bullseye fast and it is also its limit: the target says nothing about the cost of the items, their risk or their sequence.

Running the session

1. Write the goal in the centre
   One sentence, a verb and a measure, posted before the first item arrives. Have two participants read it back and compare what they understand by it.
2. Announce the primary ring's capacity
   The number of places is set before anyone has seen a placement. Announced afterwards, it becomes the subject of the discussion itself.
3. Prepare the items
   One label per item, a comparable granularity and a uniform wording, verb and object. Merge duplicates and split composite items before the session.
4. First placement round, timed
   Each participant places their items without comment or in turn. The round runs five to ten minutes, which captures judgement before the room aligns on the first opinion voiced.
5. Resolve the primary ring's overload
   The group compares the primary ring's candidates two by two against the goal and pushes back to the secondary ring whatever contributes through an intermediary. The question that unblocks it is "if we kept only this one, would the goal be met?"
6. Record the placement
   Photograph the target and write one line per item, its ring and the reason retained. A photograph alone gives back the placement without its reason.

What makes the exercise fail

The soft goal is the dominant pitfall. A centre labelled "modernise the service" makes every item defensible in a primary position, and the session produces a target full in the middle and empty at the edge, with the feeling of work done. The remedy is to reword the goal until a plausible item from the portfolio lands in the tertiary ring. As long as that test fails, the target sorts nothing.

Uneven item granularity is the second. A label carrying a whole programme crushes a label carrying one feature: the first wins the primary ring by its size and not by its proximity, and it takes in everything it contains. The split is made before the session, at a level where each item can be judged on its own.

The third pitfall turns on who places the label. A placement made under the line manager's eye records one opinion three times rather than three opinions. The deliverable is then one person's opinion. So the first round is played in silence or simultaneously, and the discussion then opens on the placement gaps: two people who put the same item two rings apart are not reading the goal the same way.

The fourth pitfall shows up after the session. The tertiary ring is read as a bin, its items drop out of follow-up and six months later the group re-examines an idea already weighed. The written record, with the reason for the placement, tells an item placed far from the centre apart from an item discarded.

AI considerations

Preparing the items is the surest gain. A language model puts a mixed list into one format, one wording per item, flags the duplicates and spots the composite labels, the ones holding an "and" or covering three separate pieces of work. Since uneven granularity distorts the placement, this formatting is worth its cost on a portfolio of thirty candidates. The same model serves to test the central goal: asking it for three readings and finding that they diverge is learning before the session that the goal is too loose.

After the session, the model drafts the record from the annotated photograph and groups the reasons given into families, which brings out the criteria the group actually applied. Where two groups have worked the same list against the same goal, it compares the two targets and names the items that changed ring, the raw material of the session that reconciles them.

What the model must not produce is the placement. A pre-filled target anchors the room on a first answer that the group then corrects at the margins: a placement without the people who made it is one more opinion, stripped of the commitment a decision taken together produces. The item list also often belongs to an unpublished roadmap, so handing it to an external service is a data-processing decision to be taken before the workshop.

Examples

The vehicle and navigation office (Service des automobiles et de la navigation) of a French-speaking canton is scoping its digital portfolio. In the centre of the target: "halve counter visits by the end of 2027". Seven candidate initiatives, three places in the primary ring.

The first round placed five initiatives in the primary ring, the SMS reminder of the inspection due date in the secondary ring and the site's visual redesign in the tertiary ring. Two places are missing, and the trade-off takes the five candidates one by one.

Pre-filling forms from the existing vehicle file: if the office kept only that one, would the goal be met? No, because a pre-filled form saves a trip only when the service can be completed online. It moves to the secondary ring. Same verdict for training counter staff in digital guidance: it steers the users who turn up anyway towards the online services, without removing a visit on its own.

Three services stay in the primary ring, the ones that remove a trip on their own: online appointment booking for the inspection, payment of the cantonal vehicle tax by eBill and ordering number plates with file tracking. The site's visual redesign stays in the tertiary ring: no counter visit is explained by the look of the site, which leaves it defensible under another goal. The SMS reminder holds in the secondary ring, since it brings the user back to an online service without being one.

Visualisations

The target is the only support that shows what the technique judges. A table files the items into a "ring" column and loses what the radial placement gives: the gap between two items is visible without reading a value and the places left in the primary ring are counted by eye. That is also what sets the bullseye apart from its neighbours visually, where ranking would line up an ordered queue and grouping would fill four labelled bands.

Two states are better than one. The same target drawn before and after the trade-off shows the whole mechanism: a primary ring overloaded with five labels, then the same ring at three with the two items pushed out one step. A target already resolved leaves the capacity constraint to be guessed at. The written record fits in a three-column table.

Cost

Tools

The original medium is physical and remains the most effective in the room: three circles drawn in marker on a whiteboard or an A0 poster, repositionable labels, one pen per participant. Drawing the primary ring narrow is part of the method. A set of coloured stickers tells rounds or stakeholder groups apart on the same target.

For a distributed team, the target fits on a shared canvas, Miro, Mural, FigJam or a plain slide edited simultaneously, changing nothing in a method that assumes no particular software. Two settings count. Lock the canvas between rounds, failing which the items move in silence and the trace of the first round is lost. Export an image at the end of each round to keep both states. The spreadsheet holds the record, where it can be sorted and compared from one session to the next.

Sources

• LUMA Institute, Innovating for People: Handbook of Human-Centered Design Methods, 2012, the Bull's-Eye Diagramming method (Understanding skill, Patterns & Priorities group): the arrangement of three concentric rings around a central goal, the Primary, Secondary and Tertiary labels, the placement of items by the group, together with the recommendations on conduct, the primary ring's capacity fixed before the start, timed rounds and the outer ring read as a distance. LUMA names and codifies a form already circulating in design practice, and the handbook turns it into a reproducible method. The technique rests on this single source, that of a human-centred design institute.