Canonical URL: https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/ --- description: Viva! reports a GreenIce search comparison. Conversion rates differ by 0.42 percentage points, but denominators, the period and user assignment are missing. title: GreenIce compares Kimera with native search: reported rates of 1.85% and 1.43% image: https://elacelerador.es/images/social/v1/greenice-compares-kimera-native-search-reported-rates-1-85-43-59229d73f762937294ad630f473bc6df.jpg --- 1. [Home](https://elacelerador.es/en/) 2. [Conversion](https://elacelerador.es/en/conversion/) # GreenIce compares Kimera with native search: reported rates of 1.85% and 1.43% Viva! reports a GreenIce search comparison. Conversion rates differ by 0.42 percentage points, but denominators, the period and user assignment are missing. By [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 12 October 2026 at 23:09 CEST 13 min read [ Add as preferred source ](https://www.google.com/preferences/source?q=elacelerador.es&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) Share [ ](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dlinkedin%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on LinkedIn. Preview requests: 0") [ ](https://wa.me/?text=GreenIce+compares+Kimera+with+native+search%3A+reported+rates+of+1.85%25+and+1.43%25+https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dwhatsapp%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on WhatsApp. Preview requests: 0") [ ](https://t.me/share/url?url=https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dtelegram%26utm%5Fmedium%3Dsocial&text=GreenIce+compares+Kimera+with+native+search%3A+reported+rates+of+1.85%25+and+1.43%25&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Telegram. Preview requests: 0") [ ](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dfacebook%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Facebook. Preview requests: 0") ![Selection of LED light bulbs and lamps against a dark background](https://elacelerador.es/images/variants/v1/greenice-search-enters-commercial-decision-selection-led-light-bulbs-lamps-against-dark-w1280.webp) Different formats of LED lamps. Archive photograph illustrating variety within a lighting catalogue. Framing adapted. Photo: [Geoffrey A. Landis · CC BY 3.0](https://commons.wikimedia.org/wiki/File:LED%5Fbulbs.jpg?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) [GreenIce](https://elacelerador.es/en/organization/greenice/) has compared its native search with Kimera, and agency viva! reports conversion rates of 1.43% and 1.85%, respectively, for a trial whose dates and sample it does not disclose. The difference deserves attention at a lighting retailer, where finding a suitable product means resolving both aesthetic preferences and technical restrictions. It also demands a careful reading of the numbers. The case presents promising results, but its percentages and counts do not fully reconstruct the experiment or establish an independently measured economic benefit. Viva!'s case study was published June 24, 2026\. GREEN ICE, S.L., tax identifier B83253468, based in Valmojado, Toledo, is identified as the store's operator in its legal notice . Publication dates do not tell us when the trial took place. This revision preserves that distinction rather than assigning the results to an unsupported period. The data behind the story ### Agency-reported rates Experiment period not published Conversion % conversion Tap or hover for details. All values are also available in the data table. Expand notesCollapse notes **Source:** viva! No sample or assignment published. Does not establish causality. Explore the data table Scroll left or right to see all columns __Agency-reported rates (% conversion)__ | Category | Conversion | | ------------- | ----------------- | | Native search | 1.43 % conversion | | Kimera | 1.85 % conversion | ## What the published numbers say Viva! reports 150 purchases with native search and 198 with Kimera, alongside 359 and 424 add-to-cart actions. The reported mobile cart counts are 95 and 130\. These are counts from the case, without a table of assigned visitors, eligible sessions or observation days that would allow every quantity to be converted into a comparable rate. The reported conversion rates differ by 0.42 percentage points. Their relative change, calculated as 1.85 divided by 1.43 minus one, is about 29.4%. Both descriptions use the same values but answer different questions. The first expresses the difference against the rate's full denominator, while the second compares the difference with the initial rate. The purchase counts provide another comparison. Moving from 150 to 198 means 48 more purchases and a 32% increase against 150\. Moving from 359 to 424 carts means 65 more actions, about 18.1%. The agency describes these changes as 29% and 16%. Its published account does not explain why those descriptions differ from straightforward calculations using the displayed counts. Different periods, populations or definitions could sit behind individual measures, but the source does not provide a reconciliation showing whether that happened. Choosing the most favourable explanation would introduce an unsupported fact. Nor does the discrepancy alone establish that the entire intervention was invalid. It identifies the documentation needed to interpret the magnitude of the reported result accurately. The mobile comparison adds 35 cart actions, a 36.8% increase against the 95 reported for native search. That intermediate result does not establish a matching increase in mobile orders. The document does not supply a device breakdown of purchases or a denominator allowing readers to determine whether the mobile share of the observed traffic changed between variants. ## A lighting catalogue faces two different searches The commercial hypothesis makes sense without accepting every conclusion in the agency account. One shopper may know the precise specification required; another may describe an appearance and intended use. Search must return useful items for both journeys, even when the catalogue has not been written using the vocabulary of every potential customer. In lighting, a visual match resolves only part of the problem. A fixture can have the desired finish and still be unsuitable for the installation. A commercially useful answer should let a buyer check requirements that cannot be established from a photograph, such as power supply, installation or compatibility when relevant. Aesthetic relevance cannot replace the technical information in a product record. A Coveo and Arlington Research survey identifies two related but different obstacles: 31% of online shoppers surveyed in the United States and United Kingdom reported problems with choices or filters, and 30% with search or browsing. The evidence dates from December 2025 and does not measure GreenIce. For a lighting catalogue, this distinction separates finding a product from narrowing it down by installation requirements. Adding visual interpretation does not, by itself, solve both tasks. The data behind the story ### Finding and filtering remain different problems United States and United Kingdom, December 2025; frequent problems reported by online shoppers. Respondents Tap or hover for details. All values are also available in the data table. Expand notesCollapse notes **Source:** Statista 4,000 online shoppers aged 18+; online survey, 14–19 December 2025\. Multiple answers. Four of seven categories selected. The original landing confirms the base and countries; the complete PDF was not obtained. Does not measure Kimera’s effect or the technical accuracy of recommendations. Explore the data table Scroll left or right to see all columns __Finding and filtering remain different problems (%)__ | Category | Respondents | | ---------------------------- | ----------- | | Difficult choices or filters | 31% | | Difficult search or browsing | 30% | | Irrelevant recommendations | 28% | | Cannot select availability | 27% | Viva! attributes image interpretation and search-attribute enrichment to Kimera. That explanation offers a mechanism through which an item might appear for a description that does not literally match its title. It remains the supplier's explanation of its product. The available documentation does not allow an audit of the model or establish its accuracy across the retailer's entire catalogue. The distinction matters to marketing work. Finding an appropriate item sooner could shorten the distance between an acquired visit and a specific purchase consideration. A visually attractive but unsuitable match could instead increase clicks and carts without improving satisfactory purchases. Recommendation quality needs an evaluation that follows the customer beyond the first interaction with a search result. An editorial analysis could distinguish reference-code queries, technical-need queries and descriptions of visual preferences. This classification is a proposed evaluation, not a published GreenIce breakdown. It would show which problems benefit from the change and help prevent an attractive overall average from concealing deterioration in exact searches for a product that a buyer already knows. ## Assigned visitors matter as much as orders People who use search are not necessarily comparable with visitors browsing categories. Someone entering a product reference may already have a more specific intention. Comparing their purchases with those of other visitors could credit search for a difference that existed before anyone used it. The intervention should therefore be assessed within a population defined before its outcomes are observed. Interpreting viva!'s trial would require knowing how visitors were assigned and whether a person retained the same variant on later visits. It also matters whether counts cover everyone assigned, only people who searched or only those who received results. Each denominator describes a different experience and can produce a different conversion rate without any underlying contradiction. Stable assignment reduces a particular difficulty when a purchase takes several sessions. A shopper might discover a fixture on a phone and return later to check dimensions. If that journey is counted as several people, or the search changes halfway through it, purchases no longer represent two separate experiences in a straightforward way. The public account does not explain how such journeys were handled. Dividing orders by a rounded rate cannot precisely reconstruct the missing group sizes. That exercise would assume the rate's numerator matches the displayed purchase count and that its measurement unit is known. Without those conditions, an apparently precise session count would introduce data the source has not published. The absence of a denominator should remain visible rather than being filled with arithmetic assumptions. Changes in the composition of traffic can also affect results. A campaign aimed at professionals, a promotion or a greater share of returning customers can alter purchase probability. The case describes an A/B comparison but does not publish its randomisation mechanism, whether variants ran simultaneously or how such differences were controlled. The label alone does not document the design. ## What a higher rate still leaves unresolved A conversion rate describes the proportion of actions or people reaching an outcome under its chosen definition. It does not reveal order value, margin or returns. Search could make inexpensive products easier to buy and raise the purchase count while reducing average contribution. The case supplies no figures capable of confirming or ruling out that possibility. A lighting retailer can also receive multi-unit orders from customers buying for a project. Counting orders, units and buyers answers different commercial questions. Evaluating the change's value would require keeping these observation units separate and examining each order's net outcome once the return period chosen for analysis has ended. More transactions do not automatically mean greater economic value. Keeping the purchase requires a separate reading. In Cetelem’s July 2024 survey of online returns in Spain, changing size, colour or model appeared in 48% of responses from the subgroup that returned products; 28% cited faults and 17% receiving a different item. These are reasons across product categories, not lighting incidents. They distinguish suitability, product condition and fulfilment errors. A visual match addresses only part of that chain, and the survey attributes none of these reasons to a search engine. The data behind the story ### Returns distinguish suitability, faults and the wrong product Spain, July 2024; selected reasons among people who returned online purchases. Respondents Tap or hover for details. All values are also available in the data table. Expand notesCollapse notes **Source:** Statista Cetelem / CanalSondeo, CAWI; overall study of 2,200 people aged 18–65, not this chart’s base. The returner subgroup count is not separately disclosed. Multiple answers covering the preceding twelve months. Table and methodology verified in Statista; original PDF not recovered. Not rates per order, GreenIce customers or causal search evidence. Explore the data table Scroll left or right to see all columns __Returns distinguish suitability, faults and the wrong product (%)__ | Category | Respondents | | ------------------------------- | ----------- | | Change of size, colour or model | 48% | | Faulty product | 28% | | Different or wrong product | 17% | The solution's costs belong in the same assessment. Licences, integration, catalogue maintenance and recommendation review use resources even when they do not appear in the conversion dashboard. A defensible economic comparison would set observed additional contribution against the additional cost of running the new experience. Two isolated percentages cannot establish that return on investment. The effect on customer support needs a separate observation. Useful recommendations can answer questions, while ambiguous recommendations can move questions to the service team. Understanding the difference would require linking service contacts to their reasons and the products bought. That is a proposed assessment criterion, rather than a service-cost reduction established by the source's account of the experiment. The persistence of any advantage as the assortment changes is also unknown. A trial using a stable catalogue may describe a different situation from a retailer adding new products or removing references. Search evaluation should consider actual availability and the currency of product attributes. Quickly finding an item that can no longer be purchased does not resolve the customer's conversion problem. ## Technical observation begins before the commercial dashboard A search journey can be separated into a concrete sequence: query, available results, selection, product record, cart and purchase. Observing each transition would help locate the obstacle. A query returning no results points to a different problem from a viewed product page that generates no purchases. A single success percentage can conceal where the tool affected the journey. That sequence requires related events without counting the same action twice. Sending an add-to-cart notification twice does not represent two customer decisions. Similarly, recording a purchase before payment confirmation may not identify a commercially valid order. Those definitions should be documented before event counts are used to decide whether a project deserves a larger investment. Response speed deserves parallel observation. Relevant search that takes too long can lose usefulness on a small screen or slow connection. A proposed evaluation would compare real response times and abandonment within the same assigned population. The case provides no published measurements allowing us to say whether Kimera improved or worsened these measures at GreenIce. Query records should not automatically become permanent personal histories. Grouped queries may be sufficient to identify a catalogue problem without permanently linking them to individuals. If a business decides to use searches to personalise later communications, the provenance, scope and legitimacy of that separate use of data need their own assessment. Availability of a signal is not permission to use it everywhere. A particular difficulty in visual interpretation is distinguishing an item from its photographic surroundings. A product may appear alongside furniture or accessories that are not part of the reference being sold. The proposed evaluation should check that search does not attribute properties of the setting to the product itself. The image is a discovery signal; verified item specifications must retain their commercial authority. Product variants also need review. A photograph showing a finish does not establish that the finish is available in every product combination. A result should lead to an item that can actually be bought and explain which attributes apply to that variant. That distinction matters because a cart can register interest before a shopper discovers an incompatibility or finds that the required combination is unavailable. A catalogue audit should record these issues by product family and relate them to the queries that produced them. This would distinguish classification problems from incomplete product records or incorrectly assigned images. The proposal does not assume that these failures exist at GreenIce. It describes the evidence needed to evaluate visual search with the standards appropriate for a technical purchase, where selecting the wrong item can defeat an apparently successful shopping journey. Marketing and merchandising also need to identify who can change result priorities. Promoting an item because of margin or surplus stock introduces a different objective from answering the query. A system can balance these decisions, but evaluation requires a record of the rules that applied during the comparison and whether they changed partway through its observation period. ## Using the case without turning it into a promise A company considering similar work can begin with queries that currently fail to find a suitable product. That review should distinguish unavailable stock, unfamiliar wording and incomplete product records. Buying new search to solve an inventory problem is a different decision from improving query interpretation and requires a different justification. The existing problem should define the proposed investment. The team can then prepare representative queries with acceptable products and explicit restrictions. This is a proposed way to assess relevance before exposing customers to a new experience. It would separate visual successes from technical errors and identify whether the system produces a plausible answer when no suitable match exists. A confident response is not necessarily a useful retail response. The commercial decision needs an experiment with a primary question fixed before it begins. One possible question is whether the new experience increases net contribution per eligible visitor. Cart actions would provide intermediate explanations. Establishing that hierarchy prevents success being declared through a secondary measure when the business outcome remains unchanged or worsens during the same period. The observation window should cover the normal purchasing journey and apply comparable return rules. Ending an evaluation as soon as a favourable difference appears does not establish that it will persist. A device analysis should also be specified beforehand when commercially relevant, rather than searching afterwards for the combination of segments displaying the largest improvement. A result consistent with the hypothesis can justify a broader evaluation rather than automatic extrapolation to the whole catalogue. Expansion should consider how much traffic actually used search and which queries account for the effect. The value of improving a particular search experience cannot directly establish an equivalent percentage increase in the company's total sales. ## What is established and what remains open The published material establishes that the agency reports two conversion rates and several counts favourable to Kimera. It permits an arithmetic check and examination of a plausible product-discovery mechanism. It does not supply the experimental file, complete assignment data, duration, uncertainty intervals or an income statement turning the case into independently demonstrated evidence of profitability. The chart therefore preserves the rates as they appear in the source and labels them as agency-reported values. Its scale expresses conversion percentages without combining them with order or cart counts. Changes in the counts are explained through reproducible calculations in the text, while differences from the promotional percentages remain visible instead of being resolved through speculation. The mechanism described by viva! uses images to expand searchable attributes. In lighting, that expansion brings the language of an appearance closer to the catalogue, but visual similarity does not establish installation compatibility. A result matching the desired finish must still allow the buyer to identify the product and consult its specification. This distinction connects discovery quality with the purchase that survives after delivery: selecting an attractive object and selecting a usable object are different decisions that the same cart action does not distinguish. For GreenIce, the concrete issue is turning a description of a need into a technically useful lighting selection. The figures justify investigating that conversion surface and seeking better documentation. The next decisive evidence would be a reproducible comparison connecting result relevance with net purchases, costs and satisfaction after product delivery, under the same measurement definitions throughout the observation period. ## In this article ### Organizations * [GreenIce](https://elacelerador.es/en/organization/greenice/) ### Sources and documents * [viva!: GreenIce](https://vivaconversion.com/blog/cro/buscador-inteligente-mejora-conversion-ecommerce-caso-greenice-con-kimera/?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · viva! * [GreenIce: identidad](https://greenice.com/policies/legal-notice?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · GreenIce * [GreenIce: compañía](https://greenice.com/pages/quienes-somos?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) · GreenIce * Statista Share [ ](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dlinkedin%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on LinkedIn. Preview requests: 0") [ ](https://wa.me/?text=GreenIce+compares+Kimera+with+native+search%3A+reported+rates+of+1.85%25+and+1.43%25+https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dwhatsapp%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on WhatsApp. Preview requests: 0") [ ](https://t.me/share/url?url=https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dtelegram%26utm%5Fmedium%3Dsocial&text=GreenIce+compares+Kimera+with+native+search%3A+reported+rates+of+1.85%25+and+1.43%25&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Telegram. Preview requests: 0") [ ](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Felacelerador.es%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F%3Futm%5Fsource%3Dfacebook%26utm%5Fmedium%3Dsocial&utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es "Share on Facebook. Preview requests: 0") ![Portrait of Marian Vasile](https://elacelerador.es/images/variants/v1/marian-vasile-autor-acelerador-retrato-w320.webp) [LinkedIn](https://www.linkedin.com/in/nairamelisav/?utm%5Fsource=elacelerador.es&utm%5Fcampaign=elacelerador.es) About the author ## [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) Marian Vasile is a MarTech and growth professional. His work spans product marketing, SEO, paid advertising and deep online tracking, including behavioral analytics and attribution. ## Related articles [Conversion](https://elacelerador.es/en/conversion/) Analysis ### [Differentiate an offer through clarity and usefulness](https://elacelerador.es/en/conversion/differentiate-an-offer-through-clarity-and-usefulness/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 13 October 2026 4 min [Conversion](https://elacelerador.es/en/conversion/) Analysis ### [Design a financially sustainable free product](https://elacelerador.es/en/conversion/design-a-financially-sustainable-free-product/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 13 October 2026 4 min [Conversion](https://elacelerador.es/en/conversion/) Analysis ### [Turn expert knowledge into a clear offer](https://elacelerador.es/en/conversion/turn-expert-knowledge-into-a-clear-offer/) [Marian Vasile](https://elacelerador.es/en/author/marian-vasile/) 13 October 2026 5 min ```json {"@context":"https://schema.org","@graph":[{"@type":["Organization","NewsMediaOrganization"],"@id":"https://elacelerador.es/#publisher","name":"El Acelerador","url":"https://elacelerador.es/","description":"Marketing and growth hacking with examples from Spanish companies.","email":"pr@elacelerador.es","publishingPrinciples":"https://elacelerador.es/en/editorial-policy/","correctionsPolicy":"https://elacelerador.es/en/editorial-policy/","verificationFactCheckingPolicy":"https://elacelerador.es/en/editorial-policy/","masthead":"https://elacelerador.es/en/about/","ownershipFundingInfo":"https://elacelerador.es/en/about/","logo":{"@type":"ImageObject","url":"https://elacelerador.es/icon-512.png?v=20261017-stacked"}},{"@type":"Person","@id":"https://elacelerador.es/author/marian-vasile/#person","name":"Marian Vasile","url":"https://elacelerador.es/en/author/marian-vasile/","description":"Marian Vasile is a MarTech and growth professional. His work spans product marketing, SEO, paid advertising and deep online tracking, including behavioral analytics and attribution.","sameAs":["https://marianvasile.com/","https://www.linkedin.com/in/nairamelisav/"],"image":{"@type":"ImageObject","url":"https://elacelerador.es/images/marian-vasile-autor-acelerador-retrato.webp","contentUrl":"https://elacelerador.es/images/marian-vasile-autor-acelerador-retrato.webp","width":519,"height":519,"caption":"Marian Vasile"}},{"@type":"ImageObject","url":"https://elacelerador.es/images/greenice-search-enters-commercial-decision-selection-led-light-bulbs-lamps-against-dark.webp","contentUrl":"https://elacelerador.es/images/greenice-search-enters-commercial-decision-selection-led-light-bulbs-lamps-against-dark.webp","width":1920,"height":1092,"caption":"Different formats of LED lamps. Archive photograph illustrating variety within a lighting catalogue. Framing adapted.","@id":"https://elacelerador.es/images/greenice-search-enters-commercial-decision-selection-led-light-bulbs-lamps-against-dark.webp#image","creditText":"Geoffrey A. Landis · CC BY 3.0","isBasedOn":"https://commons.wikimedia.org/wiki/File:LED_bulbs.jpg"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-1","name":"viva!: GreenIce","url":"https://vivaconversion.com/blog/cro/buscador-inteligente-mejora-conversion-ecommerce-caso-greenice-con-kimera/","creditText":"viva!"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-2","name":"GreenIce: identidad","url":"https://greenice.com/policies/legal-notice","creditText":"GreenIce"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-3","name":"GreenIce: compañía","url":"https://greenice.com/pages/quienes-somos","creditText":"GreenIce"},{"@type":"CreativeWork","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-4","name":"Statista","url":"https://www.statista.com/"},{"@type":"Dataset","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#dataset","url":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-top","name":"The data behind the story: GreenIce compares Kimera with native search: reported rates of 1.85% and 1.43%","description":"Data accompanying the article, with the published research figures, tables, sources and methodological notes. The CSV download contains the complete article data export.","inLanguage":"en-GB","isPartOf":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article","hasPart":[{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-1-dataset"},{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-2-dataset"},{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-3-dataset"}],"isAccessibleForFree":false,"isBasedOn":["https://vivaconversion.com/blog/cro/buscador-inteligente-mejora-conversion-ecommerce-caso-greenice-con-kimera/","https://www.statista.com/"],"publisher":{"@id":"https://elacelerador.es/#publisher"},"distribution":{"@type":"DataDownload","name":"Download data (CSV)","contentUrl":"https://elacelerador.es/en/download.csv?article=%2Fen%2Fconversion%2Fgreenice-kimera-search-conversion%2F","encodingFormat":"text/csv","requiresSubscription":true,"isAccessibleForFree":false}},{"@type":"Dataset","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-1-dataset","url":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-1","name":"Agency-reported rates","description":"Experiment period not published No sample or assignment published. Does not establish causality.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#dataset"},"isAccessibleForFree":false,"publisher":{"@id":"https://elacelerador.es/#publisher"},"variableMeasured":[{"@type":"PropertyValue","name":"Conversion","unitText":"% conversion"}],"isBasedOn":["https://vivaconversion.com/blog/cro/buscador-inteligente-mejora-conversion-ecommerce-caso-greenice-con-kimera/"]},{"@type":"Dataset","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-2-dataset","url":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-2","name":"Finding and filtering remain different problems","description":"United States and United Kingdom, December 2025; frequent problems reported by online shoppers. 4,000 online shoppers aged 18+; online survey, 14–19 December 2025. Multiple answers. Four of seven categories selected. The original landing confirms the base and countries; the complete PDF was not obtained. Does not measure Kimera’s effect or the technical accuracy of recommendations.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#dataset"},"isAccessibleForFree":false,"publisher":{"@id":"https://elacelerador.es/#publisher"},"variableMeasured":[{"@type":"PropertyValue","name":"Respondents","unitText":"%"}],"isBasedOn":["https://www.statista.com/"]},{"@type":"Dataset","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-3-dataset","url":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-figure-3","name":"Returns distinguish suitability, faults and the wrong product","description":"Spain, July 2024; selected reasons among people who returned online purchases. Cetelem / CanalSondeo, CAWI; overall study of 2,200 people aged 18–65, not this chart’s base. The returner subgroup count is not separately disclosed. Multiple answers covering the preceding twelve months. Table and methodology verified in Statista; original PDF not recovered. Not rates per order, GreenIce customers or causal search evidence.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#dataset"},"isAccessibleForFree":false,"publisher":{"@id":"https://elacelerador.es/#publisher"},"variableMeasured":[{"@type":"PropertyValue","name":"Respondents","unitText":"%"}],"isBasedOn":["https://www.statista.com/"]},{"@type":"NewsArticle","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article","url":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/","headline":"GreenIce compares Kimera with native search: reported rates of 1.85% and 1.43%","description":"Viva! reports a GreenIce search comparison. Conversion rates differ by 0.42 percentage points, but denominators, the period and user assignment are missing.","datePublished":"2026-10-12T21:09:45.032Z","dateModified":"2026-10-13T19:10:17.803Z","inLanguage":"en-GB","mainEntityOfPage":{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#webpage"},"translationOfWork":{"@id":"https://elacelerador.es/conversion/greenice-kimera-busqueda-conversion/#article"},"isPartOf":{"@id":"https://elacelerador.es/en/#website"},"isAccessibleForFree":false,"hasPart":[{"@type":"WebPageElement","isAccessibleForFree":false,"cssSelector":".reader-restricted"}],"image":[{"@id":"https://elacelerador.es/images/greenice-search-enters-commercial-decision-selection-led-light-bulbs-lamps-against-dark.webp#image"}],"articleSection":"Conversion","keywords":["GreenIce"],"mentions":[{"@type":"Organization","@id":"https://elacelerador.es/en/organization/greenice/#entity","name":"GreenIce","url":"https://elacelerador.es/en/organization/greenice/"}],"wordCount":2733,"timeRequired":"PT13M","citation":[{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-1"},{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-2"},{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-3"},{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article-source-4"}],"author":[{"@id":"https://elacelerador.es/author/marian-vasile/#person"}],"publisher":{"@id":"https://elacelerador.es/#publisher"}},{"@type":"WebPage","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#webpage","url":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/","name":"GreenIce compares Kimera with native search: reported rates of 1.85% and 1.43%","description":"Viva! reports a GreenIce search comparison. Conversion rates differ by 0.42 percentage points, but denominators, the period and user assignment are missing.","inLanguage":"en-GB","isPartOf":{"@id":"https://elacelerador.es/en/#website"},"mainEntity":{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#article"},"breadcrumb":{"@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#breadcrumbs"},"primaryImageOfPage":{"@id":"https://elacelerador.es/images/greenice-search-enters-commercial-decision-selection-led-light-bulbs-lamps-against-dark.webp#image"}},{"@type":"BreadcrumbList","@id":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/#breadcrumbs","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://elacelerador.es/en/"},{"@type":"ListItem","position":2,"name":"Conversion","item":"https://elacelerador.es/en/conversion/"},{"@type":"ListItem","position":3,"name":"GreenIce compares Kimera with native search: reported rates of 1.85% and 1.43%","item":"https://elacelerador.es/en/conversion/greenice-kimera-search-conversion/"}]}]} ```