Skip to content
Interface language: English
Design / Signal

An empty dashboard on first entry: show where to begin

Design a first-use path from an empty dashboard to a saved collection a partner can find again.

Article language: English

A person holds a red comb against a long dark braid beside a bright window.
In this article

A dashboard promises a partner “Your collections”, then displays a vast white space and a “Create” button. Create what, from what, and why? Apparently the user must supply those answers through imagination. A friendly illustration of a box sits in the middle. At least the box knows what it is.

In the fictional Rama library, a new partner needs to prepare a set of campaign materials. The first result of their own will be a saved collection containing one uploaded file, which they can later return to. It is a modest goal, but tests the whole route: understanding the empty dashboard, starting work, saving and finding the result. Closing a welcome window is not that result.

First establish what is actually empty

Before designing a prompt to act, establish the cause of the missing content with the team. A new partner with no collections, a person without permission and a user after a failed load can see an equally empty rectangle. They need different information. “Add your first collection” after an outage may encourage someone to recreate existing work the system is temporarily unable to show.

Carbon describes no data, a state following user action and problems displaying data separately. This is useful when preparing views, but does not give an application knowledge of the reason for emptiness. Agree each message's display condition with the people responsible for the particular tool.

In Rama we design four situations. Each needs its own condition and an answer to the user's question:

  • A new account has no collections of its own. Explain what will appear here and show how to begin creating one.
  • Collections exist, but the active filter excludes them. Show no matches and a way to change the filter, preserving its current value.
  • The account cannot access the selected workspace. Explain the restriction and the agreed route to access instead of encouraging copies of materials.
  • The list failed to load. Report the failed read; the proposed retry retrieves the list, rather than repeating a previous collection save.

No matches relates to communicating filters and results (in Polish): zero matches does not mean zero assets. Also check returning from another view that remembered a filter. A user may enter a screen for the first time even though the account already contains materials. A “first visit” alone is not a sufficient condition for saying there are no collections.

Show which workspace the current view belongs to. A partner working for two brands may enter a new empty workspace and think materials from the previous one have disappeared. An organisation or campaign name helps explain the list's scope. Creating another collection blindly would deepen the problem: the first file would enter the wrong place. In the test, check that the partner recognises the selected workspace before adding material. Do not defer this check until sharing.

Waiting for a read should not pretend to be confirmed absence either. If an empty state appears briefly and a list then appears, the first instruction was a false alarm. Design the transition from loading to a result or an error. For longer operations, consider the model in progress, interruption and returning to a task (in Polish).

Some workspaces do not allow the user to add anything. A partner may be waiting for an account manager to prepare materials. A “Create” button does not solve that situation. Stack Overflow Design System distinguishes an empty state with an available action from one that needs an explanation of expectations. Rama's message should reflect the agreed process: who supplies files and how the partner learns they are ready. Do not invent a date nobody has confirmed.

From “Create” to the first file

For an account able to work independently, begin with a short explanation: “Collect campaign materials in collections. Create your first collection and add a file.” Name the button “Create collection”. This answers what belongs here and what to do now without immediately explaining every library-management method.

The next screen should continue the same action. If the button leads to a form called “New asset”, the user must guess whether they reached the right place. Names after saving need consistency too: a collection should not suddenly become a project or folder without explaining the difference. This is a content-design and concept-model task, not a reason for a longer instruction on the empty screen.

Walk through the form using information a new partner actually has. They may know the campaign name but receive an internal order number only later. If a field is required by company process, show where to obtain its value. If it is unnecessary for the first save, consider deferring it. Do not call a step simple before checking its requirements.

After saving a collection, the empty state has a different scope. A collection is no longer missing; files are missing from a particular collection. Show its name and an “Add file” action. Repeating the general welcome makes it appear that the previous step was not saved. A “Start” button that still takes completed work back to the beginning of the same route is equally unsettling.

Prepare a file for the test in advance. In a research session, a partner should not have to find a private asset merely to fill an empty view. In the real tool, upload guidance should explain supported formats and limits. If upload fails, the collection may already exist. Show the file error in the right place rather than returning the whole experience to “You have no collections yet”.

Aeneia conceptual photography

After upload, the user should recognise the outcome. Show the saved file in the new collection and information matching its actual state. “Done” alone does not explain whether the file is saved, still processing or available only in the current preview. Test returning to the list and reopening the collection. Only then is it clear whether the result survived leaving the screen.

A saved file need not yet be shared with recipients. If the library separates these actions, preserve that boundary in the first route. Show the current visibility scope where the user makes the decision. Do not suggest a test asset has reached all partners when it remains in a private collection. Conversely, if the first save immediately shares a file with the team, explain that consequence before upload.

Compare this route with a tour of buttons. Five tooltips can be displayed without producing a single file of the user's own. One action can also be explained at the right moment, allowing the partner to finish the task. The number of dismissed hints describes hint use; it does not replace checking whether the intended result exists.

A sample should show a possibility, not impersonate a partner's work

A collection may be easier to explain through an example. Consider a sample set if it fits the tool's agreed functions. Name it clearly: “Example collection”. The person should understand that the files shown were not added by their company. An attractive thumbnail carrying a real customer's name can suggest a relationship that does not exist.

Separate viewing an example from creating one's own collection. If clicking copies the sample, explain what will appear in the account afterwards. If it is only an image, do not style buttons inside it as though they work. The principles of an interactive demo (in Polish) help identify that boundary, but a demonstration does not itself confirm that a partner can add their own material.

Test data also needs a considered next step. Will the example collection appear in an export or a sharing area? Will deleting the example be unambiguous? After deletion, will it still be clear how to start one's own work? Do not fill empty space with a dummy merely to make a dashboard look mature in a presentation. The user later experiences the entire behaviour, including consequences.

If a guide can be closed, check the dashboard after skipping it too. The buttons and explanations needed for the first save must remain available. Establish whether returning to help later starts it again and whether it covers an in-progress form. A person may close a window to inspect the dashboard, not because they already understand every concept and need no further guidance.

Match assistance to the actual obstacle. Someone who does not understand collections may benefit from a short example. Someone who understands but has no file yet may need a way to return or guidance on preparing material. Someone waiting for access is diverted by an upload tutorial. These three people should not receive an identical series of screens merely because none has added a file.

Some knowledge may belong in wider onboarding to company knowledge (in Polish), such as campaign naming rules. In the dashboard, show what is needed for the current decision and a relevant link to the full instruction. Sending a person to the entire company archive without identifying the right place does not answer what to enter in a required field.

Check a smaller screen and the order of navigation through interface elements too. If the illustration fills the first phone viewport, the main explanation may be below it. A screen-reader user should reach the information and available action without hearing a series of useless empty-table headings. Assess actual behaviour in the working view; a design image alone cannot establish reading order.

Test the first route, then its repetition

Before research with partners, conduct a cognitive walkthrough of first use (in Polish). At each step record what a new user knows and what the screen shows. If the correct path requires knowing an internal button name, the design hides part of the instruction. A team that has used that name for months may not notice.

In the test, give the person a campaign-related goal without dictating clicks. Observe whether they distinguish the example from their own materials, understand required data and recognise the saved result. Record assistance too. If the facilitator explains the concept of a collection, subsequent success concerns the path after that prompt. It is valuable evidence, but differs from unaided first use.

Also test entry from an invitation or saved link. If it points to a particular collection, absence of that resource needs a different response from a new account's empty list. A 404 page that helps recover context (in Polish) helps consider the next route. Redirecting every such entry to creating a new collection can hide that the user sought an already specified asset.

Finally, return to the dashboard with the same account. It should show the existing collection in the intended place. Also test deleting the last file from a collection and deleting the last collection. In both cases the user already knows the tool, but the emptiness concerns a different level. The text should still help without claiming the account was just created. Such distinctions can inform a site or tool design with Aeneia Design. In Rama, success is a collection the partner can return to. The box can then make room for the materials.

Aeneia / Aeneia editorial

Editorial article. Source links are included in the text.

About Aeneia

Brands we have worked with

  • L’Oréal
  • Garnier
  • Netflix
  • Carrefour
  • Warner Bros. Discovery
  • Biedronka
  • HBO Max
  • Wolt
  • Żabka
  • COSRX
  • Embassy of China in Poland
  • Mixa
Aeneia / Initium

From knowledge to a decision

Tell us what you want to put in order. We will agree on where to start.

Contact Aeneia

We partner with and use solutions from:

  • Google
  • OpenAI
  • Microsoft
  • Stripe
  • Anthropic
  • Linear