Creative ideas: pick what to make next
Creative ideas lists ideas for your next creative, ranked by an Opportunity score. Open it from Create > Creative ideas. Choose an idea, a product or a blank page, and the batch screen opens.
Problem and intended user
The creative strategist (the person who plans and makes ad creatives) has more products and angles than time. Choosing which product to advertise with which message is guesswork, and a weak pairing costs a batch of images that nobody uses. The media buyer (the person who runs your paid ads) wants new creatives to rest on what already sells.
Current workflow
Without Commander, the creative strategist reads Shopify admin reports and Meta Ads Manager or Google Ads results, then keeps a spreadsheet of products and messages to try. The pairing is decided by hand, and the brief goes to a design tool or an agency.
Prerequisites
- Your Shopify product catalog has finished syncing.
- Your store has angles and ads that Commander can score. Until your first ideas exist, the tab shows the How it works page (see below). It starts with Get ideas: select it to score your angles and products. It takes a few minutes, and the button gives way to a loader that reads "Building your ideas…" until it finishes. If it could not start, a message beside the button says so and the button stays. If your store has no active plan, the button gives way to "Choose a plan to get ideas" and a View plans link. You can select Start from scratch in the meantime.
- Opening the tab and choosing an idea generate nothing. Images are generated only on the batch screen, when you select Generate creatives.
How Commander works
Commander pairs a product with an angle, or with a job to be done, and gives each pair an Opportunity score. A fixed formula, not AI, calculates the score from your own store data. An insight-based idea built on a finding can also name only a product, only an angle, or neither. It is a lead you finish on the batch screen, and it has no Opportunity score (see Kinds of ideas).
Commander composes the ideas in sets of ten: eight built on an angle or a product, and two built on a job to be done. When one kind runs out, the other fills the set. The tab shows them in that order, two full rows of cards at a time: as many cards as fit two rows at the width of your screen, and never fewer than four. The list view shows ten rows at a time.
Each set mixes the kinds of idea. When your store has them, a set holds at least one Hook idea, one Concept idea, one Angle idea and one Product idea. It also holds at least one idea whose concept has results from your own ads and one whose hook has results. One idea can count for more than one of these.
Commander shows each angle once among the performance-based and suggestion-based ideas, on the product it fits best. The same angle does not repeat on several products, so a single angle that works well does not take over the page. An insight-based idea built on a finding can still use an angle that another idea uses. When you pick angles with Angles, those angles show on more than one product. The remaining places go to the ideas with the highest score.
One product also appears in only a few ideas of a set, so each set covers several products. The ideas on a job to be done count toward this too. This limit is firm: when it conflicts with the mix above, the limit wins, and a set can then lack one kind of idea.
Once Commander has chosen the ideas of a set, it lists them from the strongest to the weakest, by Opportunity score. The rank numbers follow that order, so a lower rank number never has a lower score than a higher one on the same set.
Each angle is suggested once, on the product it fits best, so the ideas stay varied. That product is not always the one the angle was made for (see Suggestion-based below).
The five kinds of idea
The label at the top of an idea says what the idea rests on most. For an idea built on an angle or a product, Commander looks at three parts of your own ads: the angle, the concept and the hook. It names the idea after the part with the strongest results. A part counts only when its ads returned more than they cost and have spent enough to prove it. A part with a small amount of spend does not name an idea, however well it did.
| Label | What it means |
|---|---|
| Angle | The angle's own results from your ads are the strongest part. |
| Concept | The results of ads with the idea's concept, the kind of creative it uses, are the strongest part. |
| Hook | The results of ads with the idea's hook, the opening line or picture that catches attention, are the strongest part. |
| Product | None of the three has proven itself: none returned more than it cost with enough spend, or none has results yet. The idea rests on the product. |
| Customer review | The idea starts from a job your customers say, in their reviews, they want done, not from an angle. |
A part with no results of its own is not counted as proof. Point at the score to see this: that row reads "no data yet" and says it is counted like your average ad. A part with results but little spend still shows its results in the score, but it cannot name the idea.
Commander finds the concept and the hook of each of your ads in the labels it gives your ads: one concept and one hook per ad. When your team corrects a label, Commander uses the correction. The concept and the hook of an angle are the ones its labelled ads spent the most on over the last 28 days. When none of those ads spent in that time, they are the ones most of its labelled ads carry. Until Commander has labelled an angle's ads, its Concept track record and Hook track record read "no data yet".
The tab opens with the heading "Creative ideas" and the line "Pick an idea, a product or a blank page to start your next creative." While your ideas load, the tab reads "Loading your creative ideas…" above two rows of placeholder cards. After you pick products, it reads "Loading ideas for 2 products…" (or "1 product") while the new ideas load, and "Loading more ideas…" beside Load more while the next ideas load.
The How it works page
The first time you open the tab, and for as long as your first ideas are not ready, the tab shows one page: How it works, "Better inputs. Better directions.". A diagram shows the data Commander brings together and the ranked ideas it leads to. The diagram shows your own store's counts, one row per source, each as "N signals":
| Source | What is counted |
|---|---|
| Products | Your active products. |
| Existing ads | Your own ads. |
| Customer reviews | The reviews Commander has for your store. |
| Concepts | The concepts your own ads use. |
| Jobs to be done | The jobs your customers want done that Commander found in their reviews, for your active products. |
| Competitor ads | The analyzed ads of the competitors you track. |
A source your store has none of stays on the page as "0 signals", in a quieter style. While the counts load the rows show no numbers, and if they cannot be read each row shows a dash and the page still works. The button beside the heading follows your first ideas:
- Before any ideas exist, it reads Get ideas.
- While Commander builds them, a loader reads "Building your ideas…". The page checks every ten seconds.
- When they are ready, the loader becomes Start creating. Commander never opens the ideas by itself, so you can read the page first. Select Start creating to go on.
Once you have selected Start creating, the tab opens on the ideas. The information button next to the idea count, labelled How it works, opens the same diagram again in a window.
Read an idea
Each idea shows:
- the label of the idea, such as Product, Angle, Concept or Hook, and a badge that names its kind: Performance-based, Insight-based or Suggestion-based (see Kinds of ideas). An insight-based idea whose finding rests on the competitors you track also carries a Competitor-based badge beside its kind;
- a rank and the Opportunity score (an insight-based idea that does not name both a product and an angle shows none). Rank 1 is the strongest idea of the first set, and the numbers keep counting on the sets that follow. Point at the score to read how it is calculated: one ring for each part, Product strength, Your winning angle, Concept track record and Hook track record. A full ring is a grade of 100 for that part, and the label at the start of each ring shows the grade and the weight. Point at a ring, or tap it, to see the points it adds and the result behind it. The details of an idea show the same rings;
- a title, and the lines under Why consider it. The title of an insight-based idea mostly starts with what to do, such as “Refresh”; for an idea that speaks to one audience or puts a product in front of buyers for the first time it starts with the angle or the product. A short line under it names the kind of creative (see Kinds of ideas). Its Why consider it sentence quotes the numbers of the finding. An insight-based idea also shows the finding it comes from, on a line that starts with Finding:. The finding is a link: select it to open the finding. When the finding is a gap between your ads and your competitors' ads, it adds one sentence about your competitors, for example "4 of your 6 competitors run “Everyday ritual” ads, and 3 of their ads have run for 30 days or more. You haven’t run it yet.". A suggestion-based idea adds a sentence in the same place and the line "Untested: none of the competitors you track and none of your ads run this angle.";
- the Featured product, or Choose a product when the idea names none. An insight-based idea that names an angle and no product shows Choose a product and the usual placeholder picture. One that names no angle shows Choose an angle under its title;
- the heart button, which saves the idea for your whole store. Everyone on your team sees the same saved ideas. An insight-based idea that lacks a product or an angle has no heart, so Saved only never lists it.
Select the title to open the idea in a Creative idea window. It adds the angle (Core argument, Parent angle or Proposed angle), the Selected scene and a Suggested message, with Previous idea and Next idea buttons and a Use idea button. Under Featured product, the product's picture and name open its product page in a new tab, so you can check the product before you use the idea. For an insight-based idea, the window also shows a Finding section with the finding's link, and a What to do differently list with two or three lines on how the new creative should differ. When the idea names no angle, a Scene section above that list says what to shoot, and Use idea puts it into the scene field of the batch screen, where you can change it. When the idea names no product, the product panel reads Choose a product.
In the window, Why consider it also names the idea's concept, the kind of creative it uses, and says whether the concept is System (from Commander) or Manual (one your team added). When your ads with that concept have results, it shows the concept's score out of 100 over 28 days and says what it means:
- "Your ads with this concept earned more than they cost over the last 28 days."
- "Your ads with this concept about broke even over the last 28 days."
- "Your ads with this concept earned less than they cost over the last 28 days."
Without results, it reads "No 28-day score for this concept yet." An idea built on a job to be done has no concept, and neither has an idea whose angle has no labelled ads yet.
Read the Product strength explanation
In the details of an idea built on an angle, the rings sit beside a short explanation of the Product strength grade. It opens with the group of products your product is compared with, for example "Out of your 43 selling products". Commander compares your product with the products that sold. When it cannot do that, it compares with all your active products instead, and the line then reads "active products".
Under that line you see up to five rows:
- the three inputs that added the most to the grade. A ranked input reads like "#4 in repeat buying", with its value when there is one, such as a percentage. The ranks count within the group above, and #1 is the best;
- one row marked Held back by. It is the input that cost the grade the most. The words Held back by come first in the row, so you do not need to see colour or a border to find it;
- the product's rank for revenue over the last 30 days, when revenue is known.
Average rating, likelihood to recommend and returns are judged on their own scale, not against your other products. They show as plain results, such as "4.7★ average rating" or "3% returned", with no rank.
Revenue is shown for context only. It is not part of the Product strength grade.
Select See what went into it to read how the grade adds up: how customers behave after buying, their experience with the product, and the demand for it, each compared with the same group of products. Select Hide what went into it to close it. An older idea may show no explanation until Commander scores your products again. Its rings still show the grade and the evidence line.
Read the Competitor proof explanation
When competitors you track run ads on the same angle as an idea, the rings in its details gain a Competitor proof ring. It compares how long your competitors keep the ads on this angle running with how long they keep their ads running in general: competitors keep paying for ads that work, so an ad that runs 30 days or more is a sign that the angle may work for them. The ring measures how long ads run, not how well they did.
The explanation opens with how many of your competitors have run ads on the angle, for example "3 of your 8 competitors have run ads on this angle". Under it, one row reads like "11 of their 12 ads on it ran 30+ days". It then says what share of all your competitors' ads run 30+ days, and how this angle scores out of 100. A score of 50 is average for your market; higher means competitors keep ads on this angle running longer than their other ads.
Ads that started less than 30 days ago and are still running are not counted yet, so a new angle is not marked down for ads that have had no time to prove themselves. An idea whose angle has no competitor ads that can be counted shows no Competitor proof ring, and its other rings carry the whole score.
Kinds of ideas
Every idea names its kind in a badge on its card and in its row. The kind is written in the badge, so you do not need to see its colour to find it.
- Performance-based ideas rest on your own results and, when you track competitors, on how long their ads on the angle keep running. Ideas built on a job to be done are always performance-based.
- Insight-based ideas come from the findings on Insights > Top findings (see Top findings). Commander reads your open findings in the order Top findings ranks them and turns every finding about one of your angles, ads or products into ideas for it. A finding about a customer objection or complaint also gives an idea. Insight-based ideas keep that order, so the first one comes from the highest-ranked finding that Commander can turn into an idea. Each idea shows the finding it comes from on a line that starts with Finding:. A finding names a product, an angle or both: an idea that names both is ready to use, and one that names only a product, only an angle or neither lets you choose the rest on the batch screen. A finding about an ad of yours that has no angle gives no idea, and neither does a finding from your Growth Audit. A finding that no creative can act on gives no idea either (see What an insight-based idea tells you to do). A finding you dismiss or resolve, or merge into another, takes its ideas out of the list the next time you open it.
- Suggestion-based ideas are untested experiments. Commander offers one for an angle that none of the competitors you track run and none of your ads run, on a product it fits. Nothing shows whether the angle works, so the idea may or may not work. The card says so in the line "Untested: none of the competitors you track and none of your ads run this angle." Commander can also suggest an angle for a product it was not made for, once it has checked that the angle fits that product. That card says who the angle was made for, in a line such as "Made for {product}. Commander thinks it also fits {product}." or "Made for {product} and 2 more. Commander thinks it also fits {product}." When Commander cannot load that name, the line reads "Commander thinks this angle also fits {product}." A suggestion is not a finding and has no place in Top findings.
When most of the competitors you track run an angle, some of their ads on it have run for a long time, and you run little or none of that angle, Commander also adds a finding about the gap to Top findings. Its idea is insight-based, and the finding reads like "Competitors run “Everyday ritual” ads and you run very few". The gap does not change an idea's Opportunity score: the score keeps resting on the rings, and the gap shows as the sentence about your competitors under the idea.
An insight-based idea keeps its place by the finding's rank, not by its Opportunity score. Commander keeps a few places in each set for insight-based ideas when your store has them, so such an idea can sit in a set above its score. Ideas of the other kinds fill the rest of the set, best score first.
What an insight-based idea tells you to do
Every insight-based idea says what the new creative is for. Commander writes this from the numbers of the finding by fixed rules, not with AI. Each idea has a title that mostly starts with what to do (the audience and first-ad titles start with the angle or the product), a line under the title, a Why consider it sentence with the finding's own numbers, and two or three lines on what to do differently. There are eleven kinds:
- Refresh: an ad on the angle is wearing out, as its click rate falls. The idea keeps the promise of the angle and changes the look. The title reads like Refresh “Everyday ritual” for {product}.
- Keep the hook, earn the click: an ad stops the scroll, but few people click. The idea keeps the opening hook and changes what follows it.
- Scale a winner: an angle or an ad has paid back and can take more budget. The idea is a new variation of it, which can run next to the winner.
- Speak to one audience: the angle works for one audience and not for the others. The idea speaks to that audience directly.
- First ad for a product: a product sells without enough ads. The title starts with First ad for when the product has never been advertised, with Put and ends with back in front of buyers when its ads stopped, and with Give and ends with its share when its ad spend is small next to its revenue.
- Test what competitors prove: competitors keep ads on an angle running and you run little or none of it. The idea makes the angle yours, in your brand's voice, and does not copy theirs.
- Differentiate: a competitor you track started new ads on an angle you run. The idea keeps the angle and changes the execution, so your ads do not look like theirs.
- Get a product in front of a peak: searches for a keyword of a product peak in a month ahead, and the product has no live ad and is in stock. The idea leads with that search.
- Re-promote: a product's sales fell against the same weeks last year. The idea gives it a fresh reason to buy now. A product whose sales rose gives no idea.
- Answer an objection: customers keep raising an objection and none of your ads answers it. The idea says it in their words and answers it.
- Answer a complaint: your reviews raise the same complaint again and again. The idea shows what the customer gets for it, in a real review's words.
Some findings ask for no creative, and they give no idea. This covers a finding about a poor conversion rate after the click, a finding about a product that sells without any ad linked to it, a repeat-purchase finding where the ad has not earned a profit, a keyword finding whose product already has a live ad or is out of stock, and the other kinds of finding on Top findings that are not about the look or the message of a creative.
An insight-based or suggestion-based idea can be on an angle your store does not have yet. Use idea then creates the angle for you when you start the batch. Saving such an idea with the heart also adds the angle to your store. An insight-based idea can also name an angle or a product and not the other, or neither. Use idea opens the batch screen with what the idea names, and you choose a product and an angle there, as the screen asks (Choose a product to generate images, Choose an angle to generate images), or write your own scene. The batch keeps the link to the finding as long as you keep what the idea named. The heart on these ideas names the idea and the angle it adds: "Save" followed by the idea's title, then "and add the angle" and the angle's name. If that idea is no longer available when you select Use idea or the heart, a banner reads "This idea is no longer available" with the line "We updated your ideas. Choose another one to continue." and Commander loads the ideas again. On the batch screen, the message reads "This idea is no longer available. Go back to Creative ideas and choose another."
Select See angle on an idea to open its angle in a side panel. It shows the angle's core statement, how many of your ads use it and how many delivered, and its variations. It also shows the angle's all-time Ad spend in your store currency, or "No delivery yet" when its ads have not delivered. An asterisk after the amount means some spend could not be converted to your store currency, so the real total can be higher. Select Manage angles to edit angles. If the angle cannot load, the panel reads "Couldn't load this angle". Select Try again.
Find an idea
- Select Products to open the product picker. While the list loads, it reads "Loading your products…". Tick one or more products, up to 50, and select Show ideas. The button then names the product, or says how many products you selected. Select the cross beside it to clear the choice. On a narrow screen, scroll the table sideways to see every column. In Demo Mode, a search box also narrows the list.
- Rank shows each product's place in the recommended order: highest Opportunity score first. Products with the same score follow their revenue over the last 30 days, and products with no score yet come last. A product keeps its rank when you filter or sort the list.
- Why it ranks here says what decided its place. Select it for the full reason, with the product's recent sales, its Insights quadrant and whether it has run in ads.
- Select Angles to open Choose angles. It lists your store's angles in a recommended order. Angles that have run in your ads come first, best return first. An angle that has spent little ranks toward your store's average return, so its own return counts for less until it has spent more. Angles not advertised yet follow, A to Z.
- Each angle you have advertised shows Paid back or Below break-even under its return. An angle is Paid back when its own ads returned more than your break-even return. The information button beside Return shows that return and the gross margin it comes from.
- Each angle shows Return, Spent and Ads from your own ads, all time. Ideas shows how many ideas use the angle and where the best one ranks. An angle no idea uses yet shows 1 idea with no rank when you have not picked products or Saved only: the idea Commander suggests for it (see below).
- Why it ranks here says what decided its place. Select it for the full reason, and select the track record to see its rings.
- The Show filters narrow the list, for example to Paid back or Used by ideas. Search and Tension narrow it too, and Sort changes the order.
- Tick the angles you want, up to 50, and select Show N ideas. Each picked angle gets its own chip with a cross, and the ideas show every idea of the picked angles. Manage angles, at the bottom, opens the page where you edit them.
- Use Idea source to choose All, Performance-based, Insight-based or Suggestion-based. All is the default and shows every kind. Performance-based shows only the ideas that rest on your own results, Insight-based only the ideas built from your findings, and Suggestion-based only the untested experiments. A chip reads "Idea source:" followed by the kind while one is on, for example "Idea source: Insight-based", and Clear filters removes it. While you filter by source, Give me more is hidden, and the list ends with a line such as "That’s every insight-based idea for now".
- Evidence is a dropdown on the same row as Idea source. It appears only while Idea source is Insight-based, and it names your choice after the colon, for example "Evidence: All". It narrows those ideas by the data their finding reads. All is the default and shows every insight-based idea. Own data only shows the ideas whose finding reads only your store's numbers. Competitor-based shows the ideas whose finding rests on the competitors you track, including a finding that reads your own numbers beside theirs. A chip reads "Evidence:" followed by the choice while one is on, and Clear filters removes it. Choosing another Idea source puts Evidence back to All. While Evidence is on, Give me more is hidden, as it is with Idea source.
- Select More filters to filter by Starting point (for example Concept or Hook) or Concept, or to show Saved only ideas.
- Concept lists every concept Commander can give your ads, not only the ones on screen. Commander sets these concepts when it analyses your ads, so this list has no page where you edit them.
- Commander applies Angles and Concept to every idea it holds, so the count covers all of them, not only the ideas loaded so far.
- Switch between Cards and List. The list view is a table with the columns Priority, Creative idea, Product and Why consider it.
- Select Load more to show the next ideas: two more rows of cards, or ten more rows in the list. Resizing the window does not change the ideas on screen; the next Load more uses the new width. Once every idea is on screen, Give me more asks Commander to look for more ideas than it has shown.
Start a creative batch
Choose one of three starts.
- Select Use idea on an idea. The batch screen opens with the idea's product, angle and scene already filled in. An insight-based idea that names no angle but has a scene opens as a Scene-led ad: the angle box reads Scene-led ad and "Built around your scene", and Commander builds the creatives around the idea's scene, so you can generate without picking an angle. To use an angle instead, select the box and pick one. On an insight-based idea, the batch also carries its finding and its lines on what to do differently. Every creative that batch makes, in the first round and in later rounds such as More like these, follows those lines. If you change the angle or the product, or pick more than one product, the batch is an ordinary one and does not carry the finding. If the finding was closed meanwhile, the batch starts without it.
- When you filter by products, Commander shows the ideas it already suggested for them first, then the job-to-be-done ideas for them, ranked the same way. It does not add every angle to the product. It also shows ideas that name no product. An idea built on a scene always shows under a product filter. An idea built on an angle shows when the angle was made for that product or Commander has checked that it fits. When Commander cannot rank a product (it is not active, it has no product score yet, or your store has no idea for it), a banner reads "{product} doesn't have ideas yet". Select Start with this product. With several such products, select Start with a product and pick one. You then choose the angle and the scene yourself. When you filter by one product and select Use idea on an idea that names no product, the batch screen opens with that product already chosen.
- Select Start from scratch at the top right. The batch screen opens empty.
Nothing is generated at this point. See Generate creatives on the batch screen.
Exact output
You get a ranked list, and a filled-in batch screen when you start from an idea.
| Part | What it shows |
|---|---|
| N creative ideas | The number of ideas that match your filters. |
| Rank and Opportunity score | The order of the ideas, and how strong Commander rates each one. |
| Label (Angle, Concept, Hook, Product or Customer review) | The part of your ads the idea rests on most. |
| Kind badge (Performance-based, Insight-based or Suggestion-based) | Where the idea comes from: your own results, a finding, an untested experiment, or an angle moved onto a product it was not made for. |
| Competitor-based badge | On an insight-based idea, that its finding rests on the competitors you track, with or without your own numbers beside theirs. |
| Own data badge | On an insight-based idea, that its finding rests only on your own numbers. |
| Title and the line under it | On an insight-based idea, a title that mostly starts with what to do (such as Refresh, Scale a winner or Test what competitors prove), or with the angle or the product for the audience and first-ad ideas, and a line that names the kind of creative. |
| Finding: line and competitor sentence | On an insight-based idea, the finding it comes from, as a link to the finding, and, for a gap, what your competitors run that you do not. On a suggestion-based idea, a sentence and the line that says it is untested. |
| Finding, Scene and What to do differently | In the idea window of an insight-based idea: the finding's link, what to shoot when the idea names no angle (Scene), and two or three lines on how the new creative should differ. |
| Why consider it | The reason Commander gives for the pairing. In the idea window, also the concept and what its 28-day score means. |
| Featured product | The product the idea is about, or Choose a product when an insight-based idea names none. |
| Choose angles | Your angles in the recommended order, with their results and how many ideas use them. |
| Use idea | Opens the batch screen on that idea. |
| Heart button | Saves the idea for your whole store, so Saved only can show it again. Your teammates see it too, and it stays saved when Commander refreshes the ideas. An insight-based idea that lacks a product or an angle has none. |
What to do with it
| What you see | Decision or action | Owner |
|---|---|---|
| An Insight-based idea | A finding on Top findings points at this product, this angle or both. Select the Finding: line to read the finding, check What to do differently in the idea window, then select Use idea to act on it. Where the idea shows Choose a product or Choose an angle, choose it on the batch screen. Filter by Insight-based to see them together. | creative strategist |
| A Suggestion-based idea | Nothing shows that this angle works, or that it works on this product, so test it small first. Select Use idea to try it, or filter by Suggestion-based to see them together. | creative strategist |
| A high-ranked idea with a product you want to push | Select Use idea, check the batch screen and generate. | creative strategist |
| A product with no idea | Filter by it to see the ideas that feature it and the ones that fit it. If Commander cannot rank it, select Start with this product and choose the angle yourself. | creative strategist |
| An idea you want later | Select the heart button. Later, use Saved only. | creative strategist |
| A concept whose ads earned less than they cost | Prefer ideas with another concept, or filter by a concept that earned more. | creative strategist |
| A new product to launch | Select Start from scratch. | media buyer |
Expected value and measurement
The tab can help you choose a product and a message on evidence from your own store, instead of by guess.
To check, compare the creatives you made from high-ranked ideas with those from ideas you chose by hand. Open Performance > Ad performance and read Spend and ROAS for the ads over the same dates. Other changes at the same time, such as a new offer or a budget change, can hide the effect of the idea.
Limitations and exceptions
- The Opportunity score is an estimate from your past results. It cannot see a trend that has not started. A new store with few ads has less to score from.
- While Commander scores your angles and products for the first time, the How it works page shows "Building your ideas…" in place of the button. You can select Start from scratch in the meantime.
- When a scoring run has finished and found no angle and product pairs to recommend, the tab reads "No ideas to show yet" and offers nothing to wait for. You can select Start from scratch in the meantime.
- If Commander cannot find angles in your store's reviews, a note at the top of the tab reads "We couldn't find angles in your reviews yet". You can still write your own angle when you start a creative batch.
- If the ideas cannot load, a banner reads "Couldn't load creative ideas". Select Try again, or start from scratch. If more ideas cannot load, the banner reads "Couldn't load more ideas".
- When no idea matches your filters, the page reads "No creative ideas match these filters". Select Clear filters. A product filter shows an insight-based idea that names no product only when it fits that product. When no idea is made for the product or fits it, a banner reads "No ideas feature this product yet", then "No idea is made for this product or fits it yet." and "Clear the filter to see every idea." Select Clear filter. With Saved only on and nothing saved, it reads "No saved creative ideas here yet".
- Saved only shows every idea you saved, also one Commander no longer suggests first. A saved idea does not show when its product is no longer active or its angle is no longer in use.
- If a heart cannot be saved, the heart goes back and a banner reads "Couldn't save this idea" (or "Couldn't remove this idea from Saved"). Select Try again, or Dismiss.
- The concept and hook of an idea come only from the labels Commander gives your ads. An angle whose ads have no label yet has no concept and no hook: both parts read "no data yet" and count like your average ad. Its ideas do not show under a Concept filter.
- Insight-based ideas need findings on Top findings that name a product, an angle or an ad of your store, or that raise a customer objection or complaint, and that ask for a creative. A finding about a poor conversion rate after the click, a product that sells with no ad linked to it, or a repeat-purchase result where the ad has not earned a profit gives no idea, so such a finding never shows under Insight-based. Without a finding that fits, no idea carries the Insight-based badge, and Insight-based shows "No insight-based ideas yet" with the line "Ideas appear here when a finding on Top findings is about one of your products, ads or angles, including competitors running an angle you barely run." The gap finding needs competitors you track and their analyzed ads, and it compares them with your own ad labels, so an angle your ads have not been labelled for counts as one you have not run. When your store has no analyzed competitor ads, the line is followed by Add competitors, a link to the competitor library.
- Suggestion-based ideas are offered for angles that none of the competitors you track and none of your ads run, which need competitors you track and their analyzed ads, and for an angle made for another product that Commander thinks fits one of yours. Their order is not a measure of how likely they are to work. Without any, Suggestion-based shows "No suggestion-based ideas yet" with the line "Suggestions appear here for an angle nobody runs yet, and for an angle made for another product that Commander thinks also fits one of yours."
- Evidence sorts findings by the data they read. A finding that rests on search demand or on data pooled across many stores shows under All only, not under Own data only or Competitor-based. A finding Commander has not yet classified, such as one saved before Evidence existed, also shows under All only. These findings carry no Competitor-based or Own data badge. When no idea matches the choice, the page reads "No creative ideas match these filters" with a line such as "No insight-based idea reads only your own data yet." or "No insight-based idea is competitor-based yet.", followed by "Set Evidence to All to see every insight-based idea." Select Show all evidence to bring the others back, or Clear filters to remove every filter.
- A finding you dismiss, resolve or merge takes its insight-based ideas out of the list the next time you open it. A new finding adds its ideas after Commander next refreshes your ideas.
- Ideas built on a job to be done are always performance-based.
- Angles and Concept show only ideas built on an angle or a product. Ideas built on a job to be done have neither, so these filters hide them.
- While Choose angles loads, it reads "Loading your angles…". If the list cannot load, it reads "Couldn't load your angles." Select Try again, or Manage angles. An asterisk after Spent means some spend could not be converted to your store currency, so the real total can be higher. An angle whose ads have not delivered shows a dash under Return, and the Not advertised yet filter under Show lists it.
- The recommended order of Choose angles uses your own ads only, all time. An angle that has spent little ranks toward your store's average return, whatever its own return, until it has spent more. Angles not advertised yet follow, A to Z, and ideas do not change the order.
- An angle filter shows whatever ideas exist for the angle, however few. A picked angle shows on its best products, so it can appear more than once. Once Commander checks the fit of your angles and products, it shows only on the products where the angle passed that check. When you also filter by products, it shows on the best of those products. When a picked angle has no idea yet, and you have not picked products or Saved only, Commander suggests one: the angle on the product that scores best with it. It shows after the other ideas, as a usual idea card. If Commander has no product to suggest the angle on, the page reads "No creative ideas match these filters".
- In Demo Mode the tab shows sample ideas with a sample data badge. Some sample ideas are insight-based, with a finding, and some are suggestion-based. Each preset shows nine insight-based ideas: a competitor gap, a refresh, a first ad, a competitor entering an angle, a keyword peak, a sales fall, a customer objection, a review complaint and a refresh that names no product. In Demo Mode the Finding: line is plain text, because the sample findings do not exist in your store: it shows the hint "Findings open in your own store". Idea source filters each kind, and Evidence separates the sample insight-based ideas that rest on competitors from those that read only your own numbers. Use idea opens the batch screen on sample data, and nothing is stored in your store. Hearts on sample ideas are kept in your browser until you close it, and are not saved for your store.