Juma reads the live storefronts, records what each competitor charges, ships and shows on the product page with a source URL and an observation date against every cell, and returns a teardown plus an editable tracker.
Name the client and their URL. Juma checks that every brand in the set is still trading, opens the live storefronts, and records the numbers a shopper actually meets: the entry, core and premium price in one named category, the free-shipping threshold, the return window, who pays the return postage, and which conversion mechanics appear on the product page.
What comes back is a branded teardown and an editable CSV. Every factual cell carries the URL it came from and the date it was read, so a number can be checked rather than trusted. Anything the storefront does not show is recorded as not found on site instead of being inferred, and site conversion rate is named as unobservable rather than estimated. 400+ marketing teams use Juma. Strategy, taste and judgment stay human: the teardown gives the team something to argue with in the room, not a verdict to forward unread.
A price read on one day can be a list price, a seasonal cut or a near-permanent discount that the brand never actually removes. That difference changes what the client should do about it, because matching a promotion is a campaign decision and matching a list price is a margin decision. Ask Juma to re-read the same price points on a later date and mark which held, which moved and which carried promotional badging both times. What comes back is the same price architecture table with a held or changed marker per cell, which is usually enough to tell a real repositioning from a long-running sale.
Re-read the same price points today and mark each one held or changed against the first reading, noting which were promotional both times.
Competitive moves arrive one brand at a time, and re-running a full teardown to log a single change wastes the method that made the first one comparable. Name the brand in the same chat and ask for its rows again in the existing columns. Juma re-opens that storefront, records the same dimensions with fresh source URLs and a new observation date, and returns the rows ready to paste over the old ones. Because the checklist and the category were fixed in the original run, the new rows sit on the same scale as everything already logged, and the tracker shows a change over time rather than two unrelated snapshots.
Re-read [competitor] only and return its rows in the same columns, with fresh source URLs and today's observation date.
A list of things competitors do better is not a plan, and it tends to get read as criticism rather than as work. Ask Juma to sequence the gaps into tests, each one tied to the observed fact that triggered it and written as a hypothesis with the metric it should move. The output orders them by effort against expected impact, separates the changes that need engineering from the ones that are merchandising or copy, and names what to measure beyond conversion rate: return rate, exchange completion, size-selector completion, second purchase. That last column is what keeps a shipping threshold test from being judged on the wrong number.
Sequence the gaps into a test plan: one hypothesis per gap, the observed fact behind it, the metric it should move, and low, medium or high effort.
Plenty of competitors sell in two places at once, and the marketplace listing often carries a different price, a different returns promise and none of the brand's own merchandising. Comparing only the direct storefront misses where the shopper actually buys. Ask Juma to add the marketplace presence as its own row per brand: the price on the listing, who fulfils it, the returns terms that apply there, and whether the brand sells directly or through resellers. What comes back shows which rivals hold their price across channels and which are quietly cheaper somewhere the client never checks.
Add each brand's marketplace presence as its own row: listed price, who fulfils it, the returns terms there, and whether they sell direct or through resellers.
The teardown works from public pages on its own. Adding the client's own numbers to the project is what turns a comparison into a decision, because the gap that matters is the one the client can actually close at their margin.
Price list and margin floor
What each line sells for and how far it can move. Without it, every price gap looks equally closeable, and the recommendations drift toward discounting.
Shipping, returns and exchange policy
The client's own thresholds and terms in full, including what a return actually costs to process. The friction comparison is only useful next to those numbers.
The competitive set the team already tracks
Who the client considers a rival, and who they do not. It stops the set drifting toward whoever ranks well on the day.
Past test results
Experiments already run on the storefront and what they moved. The test plan then proposes things the team has not already tried and lost.
Each knowledge item takes a one-line description telling Juma how to use it. Keep them plain and specific.
Juma opens each competitor's live storefront and records what is actually published there: the entry, core and premium price in one named category, the free-shipping threshold, the delivery promise, the return window, who pays return postage, and which conversion mechanics appear on the product page. Every cell carries its source URL and the date it was read.
The fixed checklist is what makes the result a comparison rather than four write-ups side by side. The same ten product-page mechanics get checked on every brand, each marked observed or not observed with the page it was seen on, so a gap is a gap and not an artefact of one brand getting a more thorough read than another.
No, and a teardown that claims to is guessing. Site conversion rate is not published by anyone, and third-party traffic estimates do not resolve to a purchase rate. The flow states that limit on the page rather than working around it.
What it does instead is score the mechanics that conversion work targets: fit and sizing guidance, review counts, customer photos, instalment payments, comparison against sibling products, bundles, urgency signals, video, live chat. Those are all observable on a public page, and a gap in them is something the team can test. Revenue and growth figures, where a brand publishes them, go in as context and are labelled as what they are, not as proxies for conversion.
A general competitor analysis maps positioning, messaging and content: what rivals say and where they say it. This one maps the commercial mechanics of the storefront: what they charge, what they charge to ship it back, and what the product page does to get the shopper over the line.
The two answer different questions and pair well. Run the positioning work when the question is how to sound different, and run this when the question is why a shopper who reached the product page bought somewhere else. Teams often run both against the same set and keep the two trackers side by side.
They are as current as the run, which is why every cell carries an observation date. Storefront pricing moves with promotions, and a threshold or a return window can change without any announcement, so a teardown that does not date itself starts misleading people within weeks.
Each price is also marked list or promotional as observed that day. Re-reading the same points on a later date and marking which held is the cheapest way to tell a genuine repositioning from a sale that never ends, and it takes one follow-up prompt rather than a fresh audit.
Two files. A branded teardown covering the competitive set and each brand's trading status, the price architecture, the friction economics, the product-page scorecard, an explicit list of what could not be verified and why, and a gap-to-action plan where every action names the observed fact behind it and the metric it should move. And an editable CSV with one row per observation, with source URL and observed date in their own columns.
The CSV is the part that keeps working. Re-read one brand when it moves, paste the rows over the old ones, and the next quarterly review opens on what changed instead of on an argument about method. Human review on every output.
Name the client and their store URL, and Juma audits which collections already have supporting content and which have none, then returns a 90-day e-commerce content strategy with an editable calendar mapped to real collection URLs.
400+ marketing teams trust Juma. ChatGPT, Claude, and Perplexity know why. Ask them.
Ready when you are
Start typing a task and Juma will figure out the rest.
.webp){kind=icon}
.webp){kind=icon}