Roadmap
Your experience lists tell you what you're running. Your roadmap tells you what you're running, when, and what happens to it next. It puts every experiment and personalization on a single timeline, next to the sales, launches, and freezes that shape your calendar.
That changes what you can see: two tests quietly competing over the same product page, a month with nothing scheduled, an experiment about to run straight through your biggest sale of the year. Each is obvious at a glance here, and fixable without leaving the page.
It's also where planning happens, rather than where finished plans get displayed. The ideas you haven't committed to a date sit in a panel beside the timeline, ready to be scheduled when you are.

Getting oriented
Click Roadmap in the left navigation. Three areas do the work:
The timeline is your plan: one bar per experience, spanning the dates it runs.
The events row, along the top of the timeline, is your context: the periods you have to plan around.
The Backlog panel, on the right, is everything you intend to run but haven't put a date on.
The toolbar decides what you're looking at.
Group
You want the timeline split into lanes, by campaign to review an initiative, or by site area to catch collisions.
Backlog
You want to see or work on your backlog. The number is how many experiences have no dates yet.
Search
You're hunting for one experience by name.
Filter chips
You want to narrow the view. Stage, Priority, Campaign, and Site area are always there; + Add a filter adds Goal, Type, Tags, and Implementation.
Time scale
You need to zoom out to plan a quarter, or in on a launch week.
‹ Today ›
You want to move through the calendar, or get back to now.
Choosing how far ahead to look
The time scale control sits at the right of the toolbar, beside Today. Month is where you'll spend most of your time. The others each have a job.
Quarter
You're planning ahead or presenting. You give up detail, since most experiences compress to short chips, in exchange for seeing the shape of a whole quarter, including where your coverage thins out.
Month
You're doing everyday planning. The default, and the best balance of range and detail.
Week
You're working out the next few weeks. Wide enough that full experience names fit.
Day
You're in a launch or promotion and need to know exactly what starts, pauses, or ends on which day.
Changing scale keeps your filters and grouping, so you can move between the long view and the close-up without rebuilding anything.
Moving through the calendar
The ‹ and › arrows shift the view backward and forward by a little less than a screen at a time, keeping some overlap so you don't lose your place. Today stays pinned where it belongs on the calendar, so after a few clicks it can sit off to one side or drop out of view. Today brings you straight back.

Grouping the timeline
Left ungrouped, the timeline is one flat set of lanes, with bars packed together by date. That's fine for a glance. Grouping is how you get answers.
Click Group and choose how to split it:
Campaign: one lane per campaign, so you can see whether an initiative is genuinely sequenced or just a pile of tests wearing the same label.
Site area: one lane per part of your store, the quickest way to find experiences that will interfere with each other.
None: back to the flat view.
A label column appears down the left, headed CAMPAIGN or SITE AREA, with a count on each group. Anything unassigned gathers under Unassigned or No site area, usually worth a look. Campaigns and site areas with nothing in them are left out, so the view stays about what you're actually running.
Fold away the groups you're not working on using the chevron (⌄) on the group header, and keep the one you're planning open.

Reading the timeline
Each experience is a bar across the dates it runs, coloured by how far it's got. That's the point: you can take in the state of your entire program without opening a single thing.
Scheduled
Planned, not started. This is your forward commitment, the part of your program you've actually decided on.
Live
Running now and collecting data. If it trails off into a faded tail past today, no end date is set and it will keep running until you decide otherwise.
Paused
Started, then stopped. A faded tail here means nothing is scheduled to change, which usually means a decision nobody has made yet.
Ended
Finished, and kept on the timeline so you can see what was running during any period you look back at.
Those faded tails matter more than they look. They're not forecasts. They mark experiences with no end date, and they're the fastest way to spot work that's drifting. A solid edge always means a real date you chose, so the difference between the two is the difference between a plan and a thing that happens to still be running.
Checking one without opening it
Hover over any bar. You get its full name, whether it's an experiment or a personalization, where it's got to, when it started or starts, and the part you actually wanted: what happens to it next. That's either No end date, or Pauses, Resumes, or Ends with the date.
It stops there on purpose. It doesn't replay every status change, because when you're planning, the question is what's coming, not what already happened.
Scheduling and rescheduling
You notice timing problems on the roadmap, so this is where you fix them. You can open the scheduling dialog three ways:
Hover over a bar and click the calendar button at its right edge.
Hover over a Ready card in your backlog and click Schedule.
Use Schedule in the footer of any experience's details.
What you can set depends on where the experience has got to:
Ready
A start, a pause, and an end. The start is the one that matters: it's what takes the experience live when the date arrives.
Live
A pause and an end. It's already running, so there's nothing left to start.
Paused
A resume and an end.
Ended
Nothing. An ended experience has nowhere left to go.
All of it is optional. Each option begins as a link (+ Add scheduled end, and so on) that opens a date and time picker when you click it. You can clear any of them again later.
The timeline redraws as soon as you save, and it shows you the difference: a date you set closes the bar off cleanly, while an open end keeps its faded tail. So the plan you just made is visible as a plan, and the things still running loose stay visible as those.
On a Ready experience, Schedule is the main action, because a date is the only thing standing between it and going live. On something already running, it sits quietly alongside Open experiment instead.

Working from your backlog
Experiences without dates don't appear on the timeline, but they're never out of sight. Backlog in the toolbar tells you how many you have and opens the panel.
That number alone is worth watching. Five means you're close to running out of planned work and your testing velocity is about to dip. Eighty means ideas aren't your problem; choosing between them is.
Ready versus Idea
The backlog splits in two. Ready holds experiences with their variants already built, waiting on nothing but a date. These are the ones you can schedule today. Idea holds the ones you've captured but not built.
The split isn't a judgement call, it's simply whether the variants exist. An experience sits in Idea until you define them, and moves to Ready on its own once you have. That makes the two sections a useful read on where your bottleneck sits: a fat Ready section means you're under-scheduling, and a fat Idea section means the work is stuck in build.
Group the timeline by campaign and your backlog groups the same way, with anything unassigned under No campaign.
What the cards tell you
Cards stay deliberately sparse: the name, plus priority, site area, and goal. Enough to triage, not enough to distract. A solid border means Ready and a dashed one means Idea, and the icon tells you whether you're looking at an experiment or a personalization.
Hover over a card and it offers whatever it needs next: Schedule on a Ready experience, Build on an Idea. You're never guessing what the next step is.
Capturing a new idea
Test ideas arrive in meetings, in support tickets, and in passing conversations, and most of them die there. + New experience on the Idea section header exists so they don't have to.
Choose Experiment to find a winning variant, or Personalization to tailor the experience for a particular audience. Give it a name and it's saved; the name is the only thing required. Fill in the rest now if you know it, or come back to it.
New experiences start as an Idea with no dates, so they land in your backlog rather than on your calendar. Nothing goes live because you wrote it down, which is what makes this safe to use as a scratchpad during a brainstorm.
An experiment carries a Hypothesis alongside Campaign, Priority, Impact, Effort, Site area, Goal, and Tags. A personalization is the same, except Description takes the place of Hypothesis and there's no Goal, since it isn't judged against a single metric in the same way. These are the same fields as your test list, so a score you set in either place means the same thing in both. See the Test planning guide for how to use Impact and Effort together.

Editing without leaving the plan
Click any bar or any backlog card to open its details. Everything you can edit in your test list, you can edit here.
There's no Save button. Each field commits as soon as you move off it. That sounds small until you're in a planning meeting, where it means you can re-prioritise, retag, and reschedule as the conversation happens instead of taking notes to type up afterwards. ← Back to list returns you to the backlog.
The details grow with the experience. An Idea shows properties only, because there's nothing built to show yet. A Ready one adds its variants with previews, and puts Schedule front and centre. A Live, Paused, or Ended one adds results above the properties: lift, visitors, and how settled the result is so far.
Properties stay editable at every stage, so an experiment that's been running for a week can still be retagged or moved to a campaign. Experiments report Goal lift as a percentage against their goal metric; personalizations report Incremental revenue in currency. A newly live experience starts out gathering data with no numbers at all, then fills in as traffic arrives.
Marking the dates you have to plan around
A test that runs through a two-week promotion produces a number you can't act on. So do the ones that span a launch, a migration, or a freeze. Events get those periods onto the same picture as your tests, so the collision is visible before it costs you a result.
There are two ways to add one. Click + Add event at the left of the events row, or hover along the row itself: a marker follows your cursor showing the date underneath, and clicking starts an event there. Name it, check the Starts date, and click Save event. The name is required, and the event appears on the row as you type it.
Leave it at a start date and you get a single marked day, with a line running down the timeline beneath it. That suits a launch day or the first day of a freeze. Add an end with + Add end date and you get the whole period instead, shaded behind your experience bars, which is what you want for a promotion, a peak season, or a migration. Remove the end date and it goes back to a single day.
Worth marking:
Peak sales periods like Black Friday and Cyber Monday, when traffic and intent look nothing like a normal week
Product launches and drops that change what visitors are shopping for
Code freezes, so you don't schedule something that can't ship
Promotions and site changes that move the baseline you're measuring against
To change an event, click its chip. Same dialog, filled in, with Delete if you no longer need it.
When events overlap they combine into a single chip showing a count (2 events, and so on) across the combined period. Click it to see what's inside and pick one to edit. They combine when their dates genuinely overlap, not merely when they look close together on screen. On a tight timeline a chip's name shortens to fit; hover over it for the full name.

Narrowing to what you care about
Search and filters run as one query across your timeline and your backlog at the same time. That's more useful than it sounds. Filter to Critical and you see both the critical work you've scheduled and the critical work still sitting undated. The gap between those two is usually the interesting part of the conversation.
Search
Type in Search your experiences to match on name. Bars and backlog cards narrow together, and the No date (n) count follows the results. The calendar keeps its shape while you search (the dates, the events, and today's line all stay put) so you never lose your bearings. Clear it with the × in the field.
With grouping on, results stay grouped: lanes keep their headers and only the matching bars remain inside them.
Filters
Stage, Priority, Campaign, and Site area live in the toolbar permanently. Click one for a checklist and tick as many values as you like; each tick applies straight away, with nothing to confirm. + Add a filter covers Goal, Type, Tags, and Implementation.
Applied filters become chips you can work with. Click one to reopen it with your selections intact, or use the ⊗ to drop it. Clear all appears as soon as anything is on, and resets your filters and your search together.
Stage
Idea, Ready, Scheduled, Live, Paused, Ended
Priority
Critical, High, Medium, Low, or no priority set
Campaign
Any of your campaigns, searchable by name, or no campaign
Site area
Global, Homepage, PDP, PLP, Cart, Landing page, Navigation, Other, or none set
Goal
AOV, CVR, RPV, CTR, ATC, or no goal
Type
Template, Theme, Page, Price, API, or no type
Tags
Any tag you've created, or no tags
Implementation
Implemented, Not implemented, or no record
If nothing matches anywhere, the timeline says so and offers to clear everything. If the only matches are in your backlog, the panel opens by itself so you don't miss them.

Putting it to work
A few ways teams get value out of the roadmap quickly.
Stop ruining your own results. Group by Site area and look for bars stacked in the same lane over the same dates. Two experiences aimed at the same part of your store at once will contaminate each other, and you won't be able to tell which change did what. Hover over one and use the calendar button to stagger it. Five minutes here saves a wasted test cycle.
Plan a quarter that survives your busy season. Put your peak periods, launches, and freezes in as events first. Switch to Quarter, then work through your Ready backlog and schedule into the space that's left. You end up with a plan that respects your constraints instead of one that collides with them in week three.
Turn intentions into a schedule. At the start of each planning cycle, open your backlog and work down the Ready section; those need nothing but a date. Anything still in Idea needs variants built first, which tells you where to point your design and dev time.
See whether your program is actually consistent. Switch to Quarter, clear your filters, and look at how densely the timeline is packed. Gaps are weeks you learned nothing. This is the fastest honest read on whether you're running a steady program or testing in bursts, and it's a better number to take to leadership than a test count.
Find the work that's drifting. Filter to Stage: Live, Paused and look for faded tails. An open-ended live test is easy to forget about, and a paused test with no resume date is almost always a decision nobody made. Both cost you capacity you could be using.
Show stakeholders the plan, not a spreadsheet. Group by campaign, filter to the one in question, and use the roadmap as your visual. It shows what's running, what's queued, and how it lines up against the events your stakeholders already have on their calendars.
Keep experiments and personalizations out of each other's way. Both live on the same timeline, so you can see everything changing on your storefront at once and make sure a personalization isn't quietly altering the experience of visitors in an active test.
Still need help?
Contact Shoplift support if you have questions about planning or scheduling your experiences.
Last updated
Was this helpful?