The same piece of awareness content does not read the same way in every job. Someone who authorizes payments cross-checks it against the requests they receive and finds a use for it straight away; someone working a service counter looks for that same mechanism in their own job and comes up empty, because the example is told through requests that never pass through there. Both people read the same thing and only one walked away with something to use. Should we build a plan for each person, then? No, that would be unmanageable. The useful question is how many groups the audience gets divided into and on what criteria, and that is where the org chart shows up, because it is the map we already have drawn.
Why does one plan for the whole organization deliver so little?
When a plan has to serve the entire organization, what stays in is whatever applies to any job, and that pushes the content towards the general. The problem shows up in the next step, because the general tends to be what people have already heard, so the piece gets read quickly, filed away and leaves nothing behind. The content can be very well made and still fail, because it is calibrated on an average, and each person’s actual work looks very little like that average.
This does not mean the general layer is redundant. There is a part of the organization that falls outside every targeted plan and that needs a well thought out baseline coverage, with its own logic for keeping it running. What delivers little is that baseline layer being the whole plan, because then the people with the most exposure receive exactly the same thing as those who work far from those processes.
Why is the org chart the first criterion we reach for?
Because it already exists. It comes imported from the directory along with the accounts, it updates when someone joins or leaves, and it has an advantage worth respecting, which is that it explains itself in a committee meeting. Finance and Operations receiving different things is accepted without discussion, whereas any other way of grouping has to be argued for.
That map has a limit, though, and it does not show until the plan is already running. It has to do with what a map like that can draw. The org chart traces the borders of each territory, that is, what each team works on, and the relief falls outside it, which is what each person can touch and what reaches them from outside. Within a single administration team there can be someone who enters suppliers’ bank details and someone who has never seen that screen. Both receive the same content because they share the group name, while the difference in risk between them is enormous.
And the other way round, the same task shows up in several boxes of the org chart. Approving a payment happens in Procurement, in the executive team and in every subsidiary, and the task carries the same exposure in all three places, even if the map puts those boxes far apart from one another.
Which criteria hold up a plan by audience?
There are four that cut across the org chart without depending on it, and each one answers a different question about the same person.
- What they can do with the access they hold. Anyone who moves money, changes payment details, grants an access or enters a system holding other people’s personal data is in a position where a mistake costs more. That group does not come ready-made from the directory. Part of it can be reconstructed from each system’s permissions, and the rest is known by whoever runs the process, so it has to be asked for.
- What reaches them from outside. Some jobs receive messages from strangers as a normal part of the work: CVs, invoices, customer requests, complaints, press enquiries. For those people the rule about distrusting whatever you did not expect cannot be applied as it stands, because what they did not expect is their job. What helps them is a verification criterion designed for strangers.
- Which channel the content arrives through. Part of the organization has no inbox and no stretch of time in front of a screen, and for them the usual repertoire falls apart entirely. That criterion governs the format before it governs the topic, and it has its own way of being built and measured.
- How much of the programme they have already covered. Someone who joined last month and someone who has spent four years completing everything assigned to them need different things. Repeating the fundamentals to the second person pushes them into doing the module on autopilot, which is the quietest way to lose someone who was already travelling with the programme.
All four have something in common, which is that none of them is answered by looking at a list of departments. So where are they answered? By asking whoever knows the process from the inside, who also happens to be the best placed to say whether the programme is working for them.
How do you build the calendar without multiplying the work?
Does that mean one calendar per audience? No, and building it that way is what brings a plan by audience down. If each audience gets its own plan, with its own dates, contents and follow-up, the coordination work multiplies by the number of audiences and the programme becomes unsustainable for the team running it.
The way out is to build the calendar in two layers. The common baseline sets the message of the month and the cadence and everyone receives it, because that is what keeps a shared conversation going and stops each department from ending up speaking its own language. The variations sit on top, changing three things and leaving the rest alone: the example (the same mechanism told through a request that audience recognizes, which is a piece of adaptation work with rules of its own), the channel and the depth.
With that structure, when a new audience appears it is enough to write a variation. And if in a given month an audience has no variation worth the effort, it receives the common baseline and nothing more. The calendar is still a single one, with one extra column. Dosage, meaning how much arrives and how often, has rules of its own and gets decided once for the whole programme.
What does the platform hold up, and what stays with whoever runs the programme?
It is worth separating these two clearly, because the ways of grouping we have been talking about are not all automated in the same way.
The structural ones get imported, because SMARTFENSE can mirror the organization that already exists in the Microsoft or Google directory and build groups, functional areas and hierarchical levels from it. They are static groupings, so they age at the pace of the synchronization, which does not always keep up with internal changes.
The ones based on behaviour and progress keep themselves up to date, and that is what smart groups are for. Their criteria can be combined with one another and they all measure activity inside the platform: risk scoring, the account creation date, the last login, campaign interactions and content completed within a time frame. A group of people who joined in the last month fills and empties on its own, at the pace of the new accounts.
And there is a finer level still, because the automatic programme decides person by person rather than by group. It respects a monthly budget of minutes, does not repeat what each person has already done and adjusts according to when they joined. On the content side, the catalogue for advanced users exists for the audiences that have already completed the basics and for those working in technical areas with administrative permissions, where a mistake has different consequences.
There is a limit worth stating plainly. Smart groups are built on people who have an account and activity in the platform. An audience without accounts cannot be segmented this way, and for them the unit of measurement shifts from the person to the shift, with everything that implies for traceability.
What stays with whoever runs the programme is the decision that gives all of the above its meaning. Which groupings are worth making in this organization is answered by knowing it from the inside, and the variations get written by whoever knows how each department speaks and what requests reach it. Once that decision is made, the platform executes it and keeps it current without anyone having to go back over it.
How do you know whether splitting by audience worked?
By looking at each audience separately, even though the dashboard invites you to look at the total. The overall average hides the group the split was made for, because an audience of thirty people with high exposure disappears inside an organization of a thousand. If the dashboard shows a number going up, it is worth opening it up by audience before counting it as an improvement.
And after that, by measuring the behaviour the group was built around. If an audience was separated out because it receives payment requests, the number that matters is how many of those requests get verified before they are paid. The count of completed modules measures something else, which is how much training that audience did, and it can rise without a single extra request being verified. When what improves is the training completed and not the verification, the group is well built and the content it receives inside does not aim at the behaviour it was separated out for.
A map that keeps itself current
The org chart will carry on being the first answer, and it is right that it should be the starting point, because it is the only map that comes drawn from the factory. What it does not do is work as the only one. Underneath those boxes there are people who share exposure without sharing a department, and they are the ones who make a plan by audience worth the work it costs.
If we had to start from a single criterion, I would start with access, because that is where a mistake reaches the loss with no steps in between. Answering who holds what does not take a new tool, it takes a conversation with whoever knows the process. Is the content that reaches the people who authorize a payment or grant an access today written for that?
Leave a Reply