Example
What a company report looks like
The org-level read: how much of the workspace is actually using it, what this company has learned to notice, and where that differs by team.
Same workspace as Otto's report, seen from the other end. Nobody is ranked anywhere on this page, and the only people named are the ones who published a card themselves — both of which are constraints rather than omissions, and the reason a report like this can exist at all.
An example workspace
Company report · 55 people · 18 months
Example — invented data, and a surface we haven't built yet
Is anybody actually using it
5,840
recognitions delivered
all time
62%
recognized in the last ninety days
34 of 55, against 31 the ninety before · 95% have been at some point
38
active voters
past 7 days
41%
poll response rate
of questions answered
6 min
answer speed
median, poll to vote
68%
reveal rate
of first recognitions get revealed 👀
The number worth looking at is the second one. 52 of 55 people have heard something specific and good about themselves — that is the thing the product is for, and it is a coverage figure rather than an engagement figure.
The response rate is 41%, which is not ninety and was never going to be. Ignoring a question costs nothing by design, so this number measures how many people felt like answering rather than how many complied. Read it alongside the coverage: a middling response rate that still reaches almost everyone is the loop working.
3 people have not been recognized yet.
That is a count and there is no list behind it. No screen in this product will tell you which 3 — the number is here so somebody can do something about it in general, not so anybody can be looked up.
How much of the company it reaches, quarter by quarter
20%
44%
53%
47%
56%
62%
And whether people are still answering
52%
47%
43%
39%
41%
41%
Recognitions per month
Last, and deliberately. This is the figure this category leads with, and it is the least informative of the three: it rises whenever a few people answer more often, without reaching one extra person. It is here to show the loop is running, not to show it is working.
What each of these does and doesn't measure
Recognitions delivered counts messages sent, one per vote. It is a volume figure and it says nothing about anyone's competence — the framework is explicit that raw volume measures delivery rather than merit, at every scope.
Active voters is a seven-day window, so it moves with holidays and launches. A dip in late December is a dip in December, not a decline.
Reveal rate is the share of first recognitions where somebody wanted the name enough to answer a question for it. Of everything on this tab it is the closest thing to a measure of whether people care.
The short version
- Strength: The craft here belongs to everybody. Helping, delivering and explaining are each said about two thirds or more of the workspace, and not one of them rests on a few shoulders.
- Strength: Being the person who turns up when somebody is stuck is the widest thing in this company — 44 of 55 people — and the most evenly spread of all twelve.
- Worth watching: Big Picture is the narrowest thing said here — 9 people — and it is getting narrower: 8 people in the six months to March, 5 in the six since. That direction is the part worth watching; the level on its own is mostly a fact about the question.
- Holding steady: Coverage is holding: 34 of 55 people were named in the last ninety days, against 31 in the ninety before.
What this company is already good at
The four things said about the most people here, and how much each rests on its top three. Widely held and evenly spread is the combination worth knowing about, because it is what a company relies on without noticing.
- 44 of 55🤝 First to Helptop three hold 12%
- 41 of 55📋 Executiontop three hold 13%
- 38 of 55🧠 Making It Cleartop three hold 15%
- 33 of 55🧩 Untanglingtop three hold 17%
None of those four is carried by a small group — the top three hold between 12% and a sixth of each. That is the useful part: this company has no single unblocker whose holiday stops the line, and no one person a new joiner has to find. It is also the reason the next section is worth reading, because it is not true everywhere.
And three worth taking into a meeting
Everything else on this report is evidence for these, or context around them.
1.71×
Go-to-market on aligning people, against this company
Alignment is a department here, not a skill.
Go-to-market is described as the team that gets everybody agreed 1.71× as often as the company is. Engineering — which shares 13 working rooms against their 9, more than anyone in the building — is described that way least of all. Getting Alignment has been said about 17 people in total.
The reflex reading is that engineers should communicate better, and it comes with a training budget attached. The room count rules that out — you cannot be overlooked and be the most-observed team here. The aligning is happening, and happening well — just somewhere else, which is why every cross-team decision here queues behind one team's calendar.
Name a rotating owner for getting agreement on cross-team projects, and don't let one team hold it twice running.
0.75×
Engineering on seeing the whole board, same comparison
The far view sits outside the room where the roadmap is written.
Data runs 1.84× the company on Big Picture and Go-to-market 1.26×. Engineering, two fifths of your headcount, runs 0.75×. Across the whole company it has been said about 9 people, and three of them account for 51% of it.
That concentration is a dependency rather than a shortage — it says nothing about whether this company thinks far enough ahead, and a great deal about what happens if two people leave. Where those people sit is the part you can act on.
Give the two teams described that way a standing seat in planning, rather than opening a requisition for a strategist.
41 of 55
people described for shipping, the flattest facet here
Shipping has stopped distinguishing anybody.
Execution has been said about 41 of your 55 people and its top three hold 13% — the flattest thing on the report.
Good news twice and a cost once. When three quarters of a company is known for delivery, delivery carries no information at promotion time, and any team whose contribution gets described in other words is arguing its case in a second language.
Ask what a person is described for that few others are, and require the evidence in those words.
What to keep an eye on
Two facets changed how many people they reach between the six months to March and the six since. The other ten moved by fewer than 3 people, which at this size is not movement, so they are not listed.
- 🔭 Big Picturefell from 8 → 5 people
- 🪑 Making Roomrose from 9 → 13 people
The one to watch is the first. Big Picture was the narrowest facet here already, and it has gone from 8 people to 5. Nothing about that says this company thinks less far ahead than it did — the same people may be doing the same thinking. What it says is that fewer people are being described that way, and since being described that way is how anyone becomes a candidate for the work, a bench that stops being named stops being visible, and a bench that stops being visible stops being drawn from.
The second is the good news and it is the more useful one to copy. Making Room went the other way, from 9 people to 13 — the largest widening on the report. Something changed in the last six months that made more people visible for making others belong. Nobody planned that as far as this data knows, and whatever caused it is worth finding before it stops.
The longer read, if you want it
Shipping, unsticking and explaining are said about almost everybody here. Judgment and direction are said about a handful.
The first half is a real asset and worth naming before anything else. The craft is spread so evenly that nobody owns it — there is no single unblocker whose holiday stops the line, no one person the new joiners all have to find. You have probably never had to build the machinery other companies your size build to compensate for not having that.
The second half is the structure underneath. The things said about only a few people are all the same kind of thing: where we are going, what to do when it is tense, telling someone something hard. Some of that is natural — not everyone can be setting direction, and a company where everyone was would have none. What is worth your attention is not how few, but which few, and where they sit.
Almost none of this is about your people. Most of it is about your rooms, and rooms are cheap to change.
Worth noticing which of those made you stop. That is the part you did not already know.
Four changes to the wiring
Open the room that isn't there.
Design and Data share no working channel, and the four names on a poll are drawn from exactly that. Over the last eighteen months the app put a member of one in front of a member of the other 34 times, where an even draw across the roster would have been about 1,800. They have never recognized each other, and nothing follows from that — at those odds you would expect about three. What follows is the 1,800. One channel changes who can appear in front of whom, mechanically, and it costs one click.
Rename the rooms you already have.
A channel counts for more when it is small and when its name says work is happening in it — the graph multiplies by 1.5 for names beginning proj-, team-, squad-, eng- or design-, and by 0.3 for announce, all-, fun-, random and general. So #all-design-data carries a fifth of the weight the same room carries as #proj-design-data. A five-fold change in who can see whom, for a rename.
Put new joiners in a cross-team room in week one.
A new joiner here waits about 4 weeks to be recognized at all and about 11 before anybody outside their own team names them. The first number is the graph filling in around somebody and you cannot move it much; the second is entirely a question of which rooms they were put in, which makes it an onboarding checklist line rather than anybody's task.
Fill in the department field.
268 recognitions belong to people whose department is blank, so they sit outside all five teams and outside every pair on the grid. One admin field, and it is the difference between four legible team reads and five.
Three things to know before you act on any of this
This is what gets said, not what happens. A quiet facet means fewer people came to mind when the question was asked. It is evidence about attention, and attention is worth managing — but it is not a measurement of whether the work is being done.
Nobody here can be looked up. There is no per-person view, no ranking and no export, on any plan. If you find yourself wanting one, the reason the answers are honest is that it does not exist — the moment recognition can be filtered to one manager's reports it becomes something people get measured with, and then nobody answers truthfully and there is nothing left to report on.
Your employees see this same page. Anyone in the workspace can sign in. There is no version of this with more in it, which is worth knowing before you quote from it.
How these numbers are made, and what they cannot be asked
Every weekday this app puts four names in front of each person and asks one question. Over the last eighteen months that is 56,800 name-appearances: 33,440 reached nobody because the poll went unanswered, 17,520 were read and somebody else came to mind, and 5,840 became a recognition. That last figure is what every number on this report is built from. The middle one is three of every four names on an answered poll, always, because one poll produces one recognition — not being picked is the default state of the mechanism rather than a result in it.
Names also fail to reach a ballot at all. Two people who share no working channel are never put in front of each other, and anyone Slack has not seen for a fortnight drops out of the pool until they are back — which is right for a poll about people you work with, and which quietly excludes part-timers, people on leave, and anyone whose work happens somewhere other than Slack. 9 people here have been put in front of fewer than 25 of the other 54, and there is no list behind that count.
Why no facet is ever called this company's strength from its vote total. The app asks each facet about equally often — it picks whichever one you have been asked about least — so every facet's supply of votes is capped at the same number. A company full of one quality cannot accumulate more votes for it; the votes redistribute among people instead. What is left between two totals is mostly the difference between two questions we wrote. So the report reads breadth, how concentrated a facet is, and how a team compares with the company on the same question — three quantities the cap does not touch.
A held cell is not a zero. Where fewer than 5 recognitions have crossed between two teams, the count is withheld — so one, two, three and four all look the same. One pair on the grid is stated as a true zero, and it is stated because the count really is zero rather than because the cell came back empty.
Two counts on this report, and no way to intersect them. 3 people are not yet recognized, on the Engagement tab. 9 have been put in front of fewer than 25 colleagues, on this one. Neither has a list, neither is broken out by team, and the report will not tell you whether they overlap. Two aggregates a reader can subtract from each other are a list wearing a disguise, which is why they are kept apart and why neither will ever get a filter.
An even draw is a yardstick, not a target. Where this report compares what happened against what an even draw across the roster would have produced, that baseline assumes everyone is equally likely to be sampled — which is exactly the assumption the channel map violates, on purpose, because a poll about people you actually work with is the point. It measures how far the graph departs from the roster. Closing it is not the goal.
How many people each of these reaches
Every one of these twelve was asked about equally often. What differs is how many people the answers landed on.
So this is not a chart of what this company talks about most — that would be a chart of which of our questions is easiest to answer. Out of 55 people, it is how many each thing has been said about at least once, at matched volume.
A narrow spoke is not a shortage, and the comparison between two spokes is weaker than it looks. Some of these describe almost anybody truthfully — plenty of people are the first to help. Others describe a small group in any organisation of any quality, because seeing the whole terrain before other people do is not a thing fifty-five people can all be doing at once. A company where everyone was named for setting direction would be a company with no direction. So the width of a spoke is partly a fact about the question and not about you, and nothing here says nine is few: knowing that would mean comparing this company against others answering the same twelve questions, which needs about ten workspaces and we currently have one.
What the chart is genuinely for is the two things underneath it, which do not depend on comparing one spoke against another: how concentrated a facet is inside itself, and how a team compares against this company on the same facet. Both cancel out everything above.
The question worth asking about a narrow spoke is not “is that too few?” but “who are they, and where do they sit?” The first has no answer from one company's data. The second has two, and they are on this page.
Big Picture is described about 9 people here, and three of them account for 51% of it. That is a dependency rather than a shortage: it says nothing about whether this company thinks far enough ahead, and quite a lot about what happens if two people leave. The concentration is the finding, and concentration is measured inside the facet, so the wording of the question cancels out of it.
The second answer is on the Departments tab, and it is the one that would change a decision. Those people are not spread evenly: Data and Go-to-market both run above this company on Big Picture, and Engineering — the largest team, and the one where the roadmap is written — runs below it. So the difficulty is less that few people see far, and more that the ones who are described that way are mostly not in the room when the direction gets set. That is a seating problem, and seating is something you can change this quarter.
Where the fewest people hold the most
🔭 Big Picture has been said about 9 people, and three of them account for 51% of it. Of the twelve, it is the one this company would lose by losing the fewest people.
That is a bus-factor reading and not a verdict on anybody. It does not say the other 46 people cannot think strategically — it says nobody has been described that way about them, which is a fact about what this organisation can currently see. Note what it is not saying, too: three of nine cannot hold less than a third of a facet whatever happens, so a narrow facet is pushed toward a high share by arithmetic alone. Measured against that floor this one is unexceptional. The reason to watch it is how few people it rests on, not how concentrated it looks.
What is deliberately not here. Nowhere does this report say this company is strong at one facet and weak at another from how many votes each got. Every facet is asked equally often, so that comparison would measure our copywriting rather than your company — and a short spoke above means fewer people have been named for it, never that the quality is missing or that anyone lacks it.
The chart above pools every recognition in the window rather than cutting it by month. That is the right default for reading breadth: how many people a facet has reached is a figure that only makes sense once it has had time to accumulate, and a single month of it would be mostly noise. Activity month by month is on the Engagement tab, where it belongs.
Why there is no percentile here
Nothing on this page compares your workspace to anyone else's. Partly that is honesty about the sample — there are not enough workspaces yet for a benchmark to mean anything, and a made-up one is worse than none.
But the deeper reason is that these questions are asked of whoever shares a channel with whom, which differs from company to company. Two workspaces' numbers are answers to differently-shaped questions, and dividing one by the other produces a figure with no referent.
Which teams share a room, and which don't
Every pair of the five departments, sorted by the working rooms between them. A working room is a public channel both teams are in; the graph weights a channel at one over its size, so a small project room is worth many times an all-hands one and the count is what predicts whether two teams can see each other at all.
Read down the rooms column. Recognition crosses where the rooms are and stops where they aren't, and the pairs at the bottom of this table are not the teams that think least of each other — they are the teams the machine has least often given the chance.
| Pair | Working rooms | Put in front | Recognitions |
|---|---|---|---|
| Engineering · Design | 5 | 4,180 | 268 |
| Engineering · Go-to-market | 4 | 5,020 | 312 |
| Engineering · Data | 3 | 2,390 | 141 |
| Go-to-market · Design | 2 | 1,640 | 96 |
| Go-to-market · Data | 2 | 1,090 | 58 |
| Engineering · People & Ops | 1 | 900 | 47 |
| Design · People & Ops | 1 | 410 | 19 |
| Go-to-market · People & Ops | 1 | 620 | 31 |
| Data · People & Ops | 1 | 240 | held |
| Design · Data | 0 | 34 | 0 |
The holes cluster among the smallest teams, and that is arithmetic rather than character: a small team is in few rooms, few rooms means a low weight on everybody's edges, and a low weight means a low place in the ordering the candidates are drawn from.
Held and zero are different states, on purpose. One pair here is under 5 exchanged recognitions, so the count is withheld and reads held. One is a true zero and is printed as one. There are no row totals on this table, because a held cell you can recover by subtracting from a total is not held — the rooms column has no such problem, which is another reason it is the spine here. Within-team recognition is not on the table at all; that belongs to the team read below.
At 55 people and five departments this is ten pairs and it fits on a page. A two-hundred-person workspace with a dozen departments has sixty-six, most of them under the floor and reading held — the rooms column is the one that still works at that size, because a channel count contains nobody's recognition and never needs withholding.
Offered, and what came back
Five departments, from the department field your Slack admin defined. What each shows is how much of the workspace could see this team, and — where there is enough to draw one — what got said about them. Never an individual, and never scoped to one manager's reports.
22
people
13
working rooms shared with other teams
public channels, from the pair table above
12,490
times put in front of somebody outside the team
over 18 months
Where this team differs from the company
Each spoke is the share of this team's recognition that went to that facet, against the same share for the workspace as a whole. Read it outward along one spoke at a time: where the solid line pulls away from the dashed one, this team is described differently from everybody else — and because both lines come from the same twelve questions, the gap between them is about the team rather than about our wording.
- Furthest above the company
- 🧩 Untangling 1.37×
- Furthest below the company
- 🎯 Getting Alignment 0.67×
Shares are shown per thousand recognitions, so a spoke at 100 means a tenth of this team's recognition. 2,340 recognitions sit behind this shape. A multiplier is this team's share of a facet divided by the whole company's — 1.00× is the dashed line, and a figure below it means this team is described that way less often than the company is, never that it does the thing less.
Thirteen working rooms with other teams, more than the next two departments put together, so Engineering appears on more of this workspace's ballots than anybody else. Against the company it runs above the line on Untangling and Execution, and below it on every Connection facet, with Getting Alignment the widest gap. Same twelve questions asked of both populations, so that difference is about this team rather than about our wording.
13
people
9
working rooms shared with other teams
public channels, from the pair table above
8,370
times put in front of somebody outside the team
over 18 months
Where this team differs from the company
Each spoke is the share of this team's recognition that went to that facet, against the same share for the workspace as a whole. Read it outward along one spoke at a time: where the solid line pulls away from the dashed one, this team is described differently from everybody else — and because both lines come from the same twelve questions, the gap between them is about the team rather than about our wording.
- Furthest above the company
- 🎯 Getting Alignment 1.71×
- Furthest below the company
- 🔍 Sharp Eye 0.48×
Shares are shown per thousand recognitions, so a spoke at 100 means a tenth of this team's recognition. 1,180 recognitions sit behind this shape. A multiplier is this team's share of a facet divided by the whole company's — 1.00× is the dashed line, and a figure below it means this team is described that way less often than the company is, never that it does the thing less.
Nine working rooms with other teams, four of them with Engineering and two with Design. Every Connection facet runs above the company here, led by Getting Alignment — the largest departure from the baseline anywhere except Data's Sharp Eye. It also carries Big Picture above the line, which matters more than it looks: a good part of this company's strategic bench sits in this thirteen-person team rather than in the twenty-two-person one where the roadmap gets written.
8
people
8
working rooms shared with other teams
public channels, from the pair table above
6,264
times put in front of somebody outside the team
over 18 months
Where this team differs from the company
Each spoke is the share of this team's recognition that went to that facet, against the same share for the workspace as a whole. Read it outward along one spoke at a time: where the solid line pulls away from the dashed one, this team is described differently from everybody else — and because both lines come from the same twelve questions, the gap between them is about the team rather than about our wording.
- Furthest above the company
- 🧠 Making It Clear 1.42×
- Furthest below the company
- 📋 Execution 0.68×
Shares are shown per thousand recognitions, so a spoke at 100 means a tenth of this team's recognition. 742 recognitions sit behind this shape. A multiplier is this team's share of a facet divided by the whole company's — 1.00× is the dashed line, and a figure below it means this team is described that way less often than the company is, never that it does the thing less.
Eight working rooms with other teams — five with Engineering, and none at all with Data. Design runs well above the company on Making It Clear and Always Learning, and it is the only team the floor lets us draw that runs above the line on Feedback That Lands. It sits below on all three Delivery facets. A team seen mostly by engineers, and described as the one that makes things legible and keeps learning.
6
people
6
working rooms shared with other teams
public channels, from the pair table above
3,754
times put in front of somebody outside the team
over 18 months
Where this team differs from the company
Each spoke is the share of this team's recognition that went to that facet, against the same share for the workspace as a whole. Read it outward along one spoke at a time: where the solid line pulls away from the dashed one, this team is described differently from everybody else — and because both lines come from the same twelve questions, the gap between them is about the team rather than about our wording.
- Furthest above the company
- 🔍 Sharp Eye 2.29×
- Furthest below the company
- 🧊 Level Head 0.31×
Shares are shown per thousand recognitions, so a spoke at 100 means a tenth of this team's recognition. 512 recognitions sit behind this shape. A multiplier is this team's share of a facet divided by the whole company's — 1.00× is the dashed line, and a figure below it means this team is described that way less often than the company is, never that it does the thing less.
Six working rooms with other teams, three with Engineering and none with Design. The sharpest departure on the report: Sharp Eye at more than twice the company rate, with Big Picture close behind, from six people. Both Multiplier facets sit near the bottom of the scale. Six people seen through three rooms produces one very clear reputation, and the rooms and the reputation are not independent of each other.
6
people
4
working rooms shared with other teams
public channels, from the pair table above
2,170
times put in front of somebody outside the team
over 18 months
Perception breakdown withheld
This scope is under 300 recognitions, which is the floor for a facet breakdown. Below it the shape is mostly sampling noise, and a chart drawn from noise is worse than no chart — so the participation numbers hold and the perception view waits.
It unlocks on its own. Nothing needs configuring.
Four working rooms with other teams — one each with Engineering, Go-to-market, Design and Data. Under the facet floor, so the breakdown is withheld: six people and 214 recognitions is not enough to draw a shape from without inventing one. Worth saying what the floor is hiding, though, because it is the opposite of what a blank cell suggests — this is the team the rest of the company most often describes in Multiplier terms, and the report cannot show you that. What it is waiting on is rooms, and a room is a thing somebody can open.
Rooms first, then reputation. Engineering is in 13 working rooms with other teams and People & Ops in 4. Almost everything that looks like a difference in how these teams are regarded is downstream of that, which is why the room counts are the first figure on every panel and the shape is the last.
268 recognitions are not in any of the five, because the people who received them have no department set. They are also missing from every pair in the table above, which is why three of those cells are less certain than they look. That is a blank admin field rather than a bug, and it is the cheapest thing on this whole report to fix.
How recognition is spread
95%
have been recognized
at least once
402
recognitions
the last thirty days
41
recognition chains
somebody praised, then praised somebody
Recognition here crosses teams more than it stays inside them. That is the healthiest pattern on the report. A workspace where praise only travels within a department is usually a workspace where nobody can see anyone else's work.
Overall it is not concentrated — but it is not evenly concentrated either. Across the workspace, no small group accounts for most of what gets said. Facet by facet the picture varies a great deal: the top three people hold 12% of First to Help and 51% of Big Picture. That spread is the thing to watch, and it is on the Analysis tab with the reach chart. If you were expecting a top ten, the absence of one is the finding.
Nobody votes the same way twice. Where two people have recognized each other repeatedly, it is almost always for different things on different days. People are not picking out of habit.
Who's on the wall
The only place this report names anybody. These are cards somebody chose to publish to the workspace — a switch they flipped themselves, off until they did, and revocable.
- Kira KoalaData AnalystThe Lanternfor Sharp Eye
- Otto OtterJunior DeveloperThe Sparkfor Always Learning
- Barry BearingtonBackend EngineerThe Vaultfor Level Head
- Penny PenguinHead of People OpsThe Hearthfor Making Room
- Felix FoxSenior EngineerThe Bridgefor First to Help
Five of 55, which is what opt-in looks like. There is no admin override and no bulk publish — the toggle only ever writes the row of whoever is signed in.
Why there is no most-recognized list, and what we show instead
A ranking of people by how much recognition they have is the one thing this product will not build, on any plan. It was specified once and deliberately dropped: the moment a list like that exists, being on it becomes the point, people vote for the outcome rather than the person, and the data stops describing anything.
What you get instead is the distribution — how wide the spread is, how much of the workspace it reaches, whether it crosses teams. At an org level that is the more useful read anyway: knowing your recognition is broad tells you something you can act on, and knowing who is top of a leaderboard tells you who is popular.
If somebody needs their own numbers, they already have them. Everyone can see their own profile in Slack, and the full personal version is Otto's report.
The ones that count for nothing
About one question in ten is a wildcard. They feed no facet, shape no card, and appear nowhere else on this report — this workspace has answered 584 of them, which is a slightly ridiculous amount of effort spent on snacks.
- Who makes Slack threads actually fun to read?96· 11 in 30d
- Who gives the best snack recommendations?88· 9 in 30d
- Who would you want next to you in a zombie apocalypse?81· 4 in 30d
- Who has the most infectious laugh?74· 8 in 30d
- Who would you road-trip cross-country with?68· 3 in 30d
- Who would absolutely crush it on a reality TV show?61· 6 in 30d
This is a count per question, not per person. Which jokes land is a fact about a company — and a more interesting one than who won them. This workspace answers the Slack-threads question more than any other wildcard, which tells you where it actually spends its day.
The thirty-day column is the one to watch, and not for the leader. Wildcard response holds up better than competency response in a busy month, which is most of the argument for having them: they keep the habit alive through the weeks when nobody has the energy for a serious question.
They exist because a fortnight of nothing but competence is a long fortnight.
What's real on that page, and what isn't
Real: the twelve facets and four pillars, the metrics and their names, the suppression floors, and the rules about what an org surface will and won't show. Every one of those is how the product behaves — the privacy page describes the same dashboard from the other side.
Invented: the workspace, the departments and every number. No real company's data was used, because we can't see one.
Not built yet: this report. The engagement figures are live in the workspace dashboard today; the written summary, the department shapes and the trend view are what the Business plan will cover, and they don't exist yet.