{"id":7175,"date":"2026-03-24T03:01:20","date_gmt":"2026-03-24T03:01:20","guid":{"rendered":"https:\/\/www.wizbrand.com\/tutorials\/personalization-test\/"},"modified":"2026-03-24T03:01:20","modified_gmt":"2026-03-24T03:01:20","slug":"personalization-test","status":"publish","type":"post","link":"https:\/\/www.wizbrand.com\/tutorials\/personalization-test\/","title":{"rendered":"Personalization Test: What It Is, Key Features, Benefits, Use Cases, and How It Fits in CRO"},"content":{"rendered":"\n

A Personalization Test<\/strong> is a structured experiment that evaluates whether tailoring an experience (message, content, offer, layout, timing, or journey) to a specific audience segment improves outcomes compared to a baseline. In Conversion & Measurement<\/strong>, it\u2019s not enough to \u201cpersonalize\u201d and hope for the best\u2014you need evidence that the change caused a meaningful lift, not just a nice-looking dashboard.<\/p>\n\n\n\n

For CRO<\/strong> (conversion rate optimization), a Personalization Test is the bridge between insight and impact. It turns hypotheses like \u201cnew visitors need more reassurance\u201d into measurable, decision-ready results. Done well, it reduces guesswork, protects the user experience, and helps teams scale personalization without inflating risk or complexity.<\/p>\n\n\n\n

What Is Personalization Test?<\/h2>\n\n\n\n

A Personalization Test<\/strong> is an experiment designed to quantify the incremental value of personalization versus a control experience. The \u201cpersonalization\u201d may be as simple as changing a headline for a segment or as advanced as dynamically assembling page modules based on behavior. The \u201ctest\u201d part means you control exposure, track outcomes, and decide based on data\u2014not assumptions.<\/p>\n\n\n\n

At its core, the concept is causal measurement: did personalization cause<\/em> better performance for a defined audience under defined conditions? In business terms, a Personalization Test answers questions like:<\/p>\n\n\n\n

    \n
  • Will segmenting messaging increase sign-ups without hurting downstream activation?<\/li>\n
  • Does returning-user personalization raise average order value, or just shift who buys?<\/li>\n
  • Which audience benefits from personalization, and which one doesn\u2019t?<\/li>\n<\/ul>\n\n\n\n

    Within Conversion & Measurement<\/strong>, a Personalization Test sits alongside other experimentation practices as a way to connect audience strategy to measurable conversion impact. Within CRO<\/strong>, it\u2019s one of the most powerful methods for improving relevance while staying disciplined about performance trade-offs.<\/p>\n\n\n\n

    Why Personalization Test Matters in Conversion & Measurement<\/h2>\n\n\n\n

    Personalization often \u201cfeels\u201d right, but feelings don\u2019t allocate budget. A Personalization Test<\/strong> matters because it converts personalization from a design philosophy into an accountable growth lever.<\/p>\n\n\n\n

    Strategically, it helps teams move from broad optimization (\u201cmake the page better\u201d) to targeted optimization (\u201cmake it better for the people who struggle most\u201d). That focus can be a competitive advantage: you can win in crowded markets by meeting intent more precisely, not just by spending more.<\/p>\n\n\n\n

    From a Conversion & Measurement<\/strong> standpoint, it improves decision quality by isolating what actually moved the metric. From a CRO<\/strong> standpoint, it creates a repeatable way to grow conversion rate, revenue, retention, and lead quality while managing risk through controlled rollout and statistically sound evaluation.<\/p>\n\n\n\n

    How Personalization Test Works<\/h2>\n\n\n\n

    A Personalization Test<\/strong> is practical experimentation with audience-specific experiences. While implementations vary, most follow a consistent workflow:<\/p>\n\n\n\n

      \n
    1. \n

      Input or trigger (what defines the audience and moment)<\/strong>\n You identify a segmentation rule (e.g., \u201cnew vs returning,\u201d \u201cpaid search vs organic,\u201d \u201cindustry,\u201d \u201ccart value,\u201d \u201cfeature usage,\u201d \u201clocale,\u201d \u201cstage in onboarding\u201d). You also define where personalization applies: landing page, pricing page, email, in-app prompt, or checkout.<\/p>\n<\/li>\n

    2. \n

      Analysis or processing (what insight becomes a hypothesis)<\/strong>\n You use qualitative and quantitative signals\u2014funnels, drop-offs, search intent, customer interviews, session replays, support tickets\u2014to form a hypothesis: \u201cThis segment needs X information to feel confident.\u201d<\/p>\n<\/li>\n

    3. \n

      Execution or application (how the experience changes)<\/strong>\n You create a control and one or more personalized variants. Then you randomize eligible users into control vs personalized experiences (or into multiple personalized approaches), ensuring the test design supports causal inference.<\/p>\n<\/li>\n

    4. \n

      Output or outcome (how results are measured and decided)<\/strong>\n You measure primary conversion outcomes and guardrail metrics (like bounce rate, revenue per visitor, churn, or refund rate). You evaluate uplift, segment effects, and trade-offs\u2014then decide to ship, iterate, or stop.<\/p>\n<\/li>\n<\/ol>\n\n\n\n

      In strong Conversion & Measurement<\/strong> practice, the \u201ctest\u201d is as important as the \u201cpersonalization.\u201d Without a test, you can\u2019t reliably distinguish true lift from seasonality, channel mix changes, or regression to the mean\u2014issues that commonly mislead CRO<\/strong> programs.<\/p>\n\n\n\n

      Key Components of Personalization Test<\/h2>\n\n\n\n

      A reliable Personalization Test<\/strong> requires more than a targeting rule. The essential components include:<\/p>\n\n\n\n

        \n
      • Clear hypothesis and scope<\/strong>: What exactly changes, for whom, and why it should improve behavior.<\/li>\n
      • Audience definition and eligibility<\/strong>: Rules based on consistent identifiers (cookies, login state, CRM ID) and an inclusion\/exclusion plan (e.g., exclude employees, bots, or low-quality traffic).<\/li>\n
      • Control experience<\/strong>: A stable baseline to compare against, unchanged during the test window.<\/li>\n
      • Experiment design<\/strong>: Randomization, traffic allocation, duration, and minimum detectable effect planning.<\/li>\n
      • Instrumentation<\/strong>: Correct event tracking, consistent attribution, deduplication, and a measurement plan aligned to Conversion & Measurement<\/strong> standards.<\/li>\n
      • Primary and guardrail metrics<\/strong>: One main success metric plus \u201cdo no harm\u201d metrics to protect downstream outcomes.<\/li>\n
      • Governance and roles<\/strong>: Who owns hypotheses, implementation, QA, analysis, and decision-making\u2014critical for scaling CRO<\/strong> responsibly.<\/li>\n
      • Documentation<\/strong>: Test purpose, configuration, results, and learnings so future teams don\u2019t repeat mistakes.<\/li>\n<\/ul>\n\n\n\n

        Types of Personalization Test<\/h2>\n\n\n\n

        \u201cTypes\u201d of Personalization Test<\/strong> are usually best described by where the personalization logic comes from<\/em> and how targeted the experience is<\/em>. Common distinctions include:<\/p>\n\n\n\n

        Rule-based (segment) personalization tests<\/h3>\n\n\n\n

        You define deterministic rules: \u201cIf source is email, show a loyalty message.\u201d These are easier to interpret and often best for early CRO<\/strong> experimentation.<\/p>\n\n\n\n

        Behavioral personalization tests<\/h3>\n\n\n\n

        Eligibility depends on actions: pages viewed, product categories browsed, feature usage, or time since last visit. These can outperform static segments but require careful tracking and identity handling in Conversion & Measurement<\/strong>.<\/p>\n\n\n\n

        Contextual personalization tests<\/h3>\n\n\n\n

        Personalization uses context signals like device type, locale, time, or inventory availability. These tests are useful when context materially changes user intent or constraints.<\/p>\n\n\n\n

        Lifecycle personalization tests<\/h3>\n\n\n\n

        Experiences differ by customer stage (trial, onboarding, expansion, win-back). These are especially valuable for subscription businesses where \u201cconversion\u201d includes activation and retention, not just the first purchase.<\/p>\n\n\n\n

        Single-page vs journey-level personalization tests<\/h3>\n\n\n\n

        Some tests personalize one page element; others personalize multiple touchpoints. Journey-level tests can produce larger impact but are harder to measure and attribute cleanly\u2014so stronger Conversion & Measurement<\/strong> rigor is required.<\/p>\n\n\n\n

        Real-World Examples of Personalization Test<\/h2>\n\n\n\n

        Example 1: SaaS pricing page by intent segment<\/h3>\n\n\n\n

        A B2B SaaS company runs a Personalization Test<\/strong> on its pricing page. Visitors from \u201centerprise intent\u201d keywords see a version emphasizing security, compliance, and procurement-friendly purchasing. Everyone else sees the standard benefits-led version. The team measures demo requests (primary) and trial starts (guardrail) to ensure the personalization doesn\u2019t reduce self-serve growth. This is classic CRO<\/strong> applied with disciplined Conversion & Measurement<\/strong>.<\/p>\n\n\n\n

        Example 2: Ecommerce returning visitors and cart sensitivity<\/h3>\n\n\n\n

        An ecommerce brand tests personalized free-shipping thresholds for returning visitors who previously abandoned carts. The personalized variant highlights a threshold aligned with their typical basket size and surfaces \u201ccomplete your set\u201d recommendations. The test tracks conversion rate, average order value, and refund rate. The outcome reveals that conversion rises, but margin drops for a subset\u2014prompting a revised targeting rule.<\/p>\n\n\n\n

        Example 3: Content marketing lead capture by industry<\/h3>\n\n\n\n

        A publisher with multiple verticals runs a Personalization Test<\/strong> where the lead magnet and CTA copy are tailored to the visitor\u2019s category consumption (e.g., cybersecurity vs cloud). The team evaluates lead volume and lead-to-opportunity rate to avoid \u201cmore leads, worse leads.\u201d This keeps CRO<\/strong> aligned with business value, not vanity metrics, within a Conversion & Measurement<\/strong> framework.<\/p>\n\n\n\n

        Benefits of Using Personalization Test<\/h2>\n\n\n\n

        A well-run Personalization Test<\/strong> can deliver benefits beyond a single lift:<\/p>\n\n\n\n

          \n
        • Higher conversion performance<\/strong>: Better message-to-intent fit often improves conversion rate, revenue per visitor, or qualified lead rate.<\/li>\n
        • More efficient spend<\/strong>: Personalization that improves landing performance can reduce effective acquisition costs by converting more of the traffic you already paid for.<\/li>\n
        • Improved user experience<\/strong>: Relevance reduces friction\u2014especially for users with distinct needs (new vs returning, SMB vs enterprise, novice vs expert).<\/li>\n
        • Faster learning loops<\/strong>: Tests reveal which segments are sensitive to what levers, strengthening future CRO<\/strong> hypotheses.<\/li>\n
        • Safer scaling<\/strong>: Controlled experimentation reduces the risk of rolling out personalization that harms key outcomes, a core Conversion & Measurement<\/strong> principle.<\/li>\n<\/ul>\n\n\n\n

          Challenges of Personalization Test<\/h2>\n\n\n\n

          A Personalization Test<\/strong> also introduces pitfalls that teams should plan for:<\/p>\n\n\n\n

            \n
          • Identity and tracking gaps<\/strong>: Cross-device behavior, cookie restrictions, and logged-out users can weaken targeting accuracy and measurement integrity in Conversion & Measurement<\/strong>.<\/li>\n
          • Small sample sizes<\/strong>: Segments can be too narrow to reach statistical confidence quickly, slowing CRO<\/strong> velocity.<\/li>\n
          • Confounding variables<\/strong>: Concurrent campaigns, site changes, or channel mix shifts can distort interpretation if governance is weak.<\/li>\n
          • Over-personalization<\/strong>: Highly tailored experiences can feel \u201ccreepy,\u201d reduce trust, or create inconsistent brand experiences.<\/li>\n
          • Misaligned success metrics<\/strong>: A lift in clicks might reduce revenue quality, retention, or downstream activation\u2014so guardrails must be real, not decorative.<\/li>\n
          • Operational complexity<\/strong>: More variants, rules, and audiences increase QA burden, performance risk, and content maintenance needs.<\/li>\n<\/ul>\n\n\n\n

            Best Practices for Personalization Test<\/h2>\n\n\n\n

            To make a Personalization Test<\/strong> credible and scalable, apply these practices:<\/p>\n\n\n\n

              \n
            1. \n

              Start with high-impact segments<\/strong>\n Prioritize segments with high volume and clear intent differences (e.g., new vs returning, branded vs non-branded, trial vs paid). This improves learning speed for CRO<\/strong>.<\/p>\n<\/li>\n

            2. \n

              Write hypotheses that can be falsified<\/strong>\n \u201cPersonalization will increase conversions\u201d is too vague. Use: \u201cFor first-time visitors from non-branded search, adding proof and risk reducers will increase trial starts by X% without increasing churn.\u201d<\/p>\n<\/li>\n

            3. \n

              Define one primary metric and multiple guardrails<\/strong>\n In Conversion & Measurement<\/strong>, guardrails protect long-term value: revenue per visitor, retention, complaint rate, refund rate, unsubscribe rate, or time-to-value.<\/p>\n<\/li>\n

            4. \n

              Keep variants meaningfully different<\/strong>\n Tiny differences produce noisy results. Make changes that match the hypothesized friction (clarity, trust, relevance, incentives).<\/p>\n<\/li>\n

            5. \n

              Control for collisions with other experiments<\/strong>\n Avoid overlapping tests affecting the same users and pages unless you\u2019re intentionally running a factorial design.<\/p>\n<\/li>\n

            6. \n

              Document targeting logic and QA thoroughly<\/strong>\n Ensure the right users see the right experience, consistently. A broken rule turns your Personalization Test<\/strong> into random chaos.<\/p>\n<\/li>\n

            7. \n

              Plan what happens after the test<\/strong>\n If it wins, define rollout steps, monitoring, and revalidation windows. Personalization performance can decay as audiences and competitors change\u2014so CRO<\/strong> remains continuous.<\/p>\n<\/li>\n<\/ol>\n\n\n\n

              Tools Used for Personalization Test<\/h2>\n\n\n\n

              A Personalization Test<\/strong> typically involves a stack of systems rather than one tool. In vendor-neutral terms, teams commonly use:<\/p>\n\n\n\n

                \n
              • Analytics tools<\/strong>: Event tracking, funnels, cohort analysis, and attribution supporting Conversion & Measurement<\/strong>.<\/li>\n
              • Experimentation platforms<\/strong>: Randomization, variant delivery, traffic allocation, and statistical reporting for CRO<\/strong> experiments.<\/li>\n
              • Tag management systems<\/strong>: Consistent deployment of tracking and experimentation scripts with version control.<\/li>\n
              • Customer data platforms \/ data pipelines<\/strong>: Unified profiles and segment definitions across web, app, and email.<\/li>\n
              • CRM systems<\/strong>: Audience attributes (industry, lifecycle stage) and downstream outcome tracking (lead quality, pipeline).<\/li>\n
              • Marketing automation tools<\/strong>: Email, lifecycle messaging, and journey orchestration tied to segment logic.<\/li>\n
              • Reporting dashboards \/ BI tools<\/strong>: Blending experiment exposure with revenue, retention, and operational metrics.<\/li>\n
              • UX research tools<\/strong>: Heatmaps, surveys, and session analysis to generate better hypotheses for Personalization Test ideas.<\/li>\n<\/ul>\n\n\n\n

                Metrics Related to Personalization Test<\/h2>\n\n\n\n

                Because personalization can shift multiple parts of the funnel, a strong Personalization Test<\/strong> uses layered measurement:<\/p>\n\n\n\n

                Primary conversion metrics<\/h3>\n\n\n\n
                  \n
                • Conversion rate (purchase, sign-up, demo request)<\/li>\n
                • Revenue per visitor \/ per session<\/li>\n
                • Qualified lead rate (where qualification is defined operationally)<\/li>\n<\/ul>\n\n\n\n

                  Secondary and guardrail metrics<\/h3>\n\n\n\n
                    \n
                  • Average order value, margin per visitor<\/li>\n
                  • Activation rate, time-to-value, feature adoption (for SaaS)<\/li>\n
                  • Retention, churn, renewal rate<\/li>\n
                  • Bounce rate, engagement depth, scroll, repeat visits<\/li>\n
                  • Refunds, chargebacks, support contact rate (quality signals)<\/li>\n<\/ul>\n\n\n\n

                    Experiment quality metrics<\/h3>\n\n\n\n
                      \n
                    • Sample ratio mismatch checks (traffic split sanity)<\/li>\n
                    • Eligibility rate (how many users truly qualify)<\/li>\n
                    • Exposure and contamination rate (users seeing multiple variants)<\/li>\n
                    • Performance impact (page speed, error rates), which can affect Conversion & Measurement<\/strong> outcomes independently<\/li>\n<\/ul>\n\n\n\n

                      Future Trends of Personalization Test<\/h2>\n\n\n\n

                      The future of Personalization Test<\/strong> is shaped by automation, privacy, and better causal measurement.<\/p>\n\n\n\n

                        \n
                      • More automation in variant creation and targeting<\/strong>: Teams will increasingly generate and adapt content at scale, but testing will remain necessary to prevent \u201cautomated mediocrity\u201d from spreading.<\/li>\n
                      • Shift from individual tracking to privacy-aware approaches<\/strong>: Less reliance on third-party identifiers means more emphasis on first-party data, contextual signals, and on-site behavior\u2014changing how Conversion & Measurement<\/strong> is implemented.<\/li>\n
                      • Incrementality and experimentation discipline<\/strong>: As attribution gets noisier, controlled experiments become even more valuable for proving true lift, reinforcing the centrality of CRO<\/strong> methods.<\/li>\n
                      • Journey experimentation<\/strong>: More tests will span multiple touchpoints (web + email + product). This raises measurement complexity, pushing teams toward stronger governance and unified measurement models.<\/li>\n
                      • Personalization quality and brand consistency<\/strong>: Future Personalization Test<\/strong> programs will evaluate not only conversion but also trust, clarity, and long-term brand impact.<\/li>\n<\/ul>\n\n\n\n

                        Personalization Test vs Related Terms<\/h2>\n\n\n\n

                        Personalization Test vs A\/B testing<\/h3>\n\n\n\n

                        A\/B testing is a broad method for comparing two or more variants. A Personalization Test<\/strong> is a specific application of A\/B testing (or experimentation) where the variant is tailored to a segment or context. In practice, personalization experiments often require additional segmentation logic and deeper Conversion & Measurement<\/strong> guardrails.<\/p>\n\n\n\n

                        Personalization Test vs segmentation<\/h3>\n\n\n\n

                        Segmentation is the act of dividing an audience into groups based on attributes or behavior. A Personalization Test<\/strong> evaluates whether using that segmentation to change the experience produces incremental improvement. Segmentation is the \u201cwho\u201d; the test proves whether tailoring for that \u201cwho\u201d works.<\/p>\n\n\n\n

                        Personalization Test vs recommendation engines<\/h3>\n\n\n\n

                        Recommendation engines generate product or content suggestions using algorithms. A Personalization Test<\/strong> can validate whether recommendations help (and which recommendation logic is best) versus a non-personalized baseline, especially within CRO<\/strong> programs focused on revenue per visitor.<\/p>\n\n\n\n

                        Who Should Learn Personalization Test<\/h2>\n\n\n\n
                          \n
                        • Marketers<\/strong> benefit by improving relevance across paid, email, and landing experiences while proving impact in Conversion & Measurement<\/strong>.<\/li>\n
                        • Analysts<\/strong> gain a rigorous framework for causal inference, guardrail design, and trustworthy reporting.<\/li>\n
                        • Agencies<\/strong> can offer higher-value experimentation programs that go beyond \u201cbest practices\u201d into measurable growth outcomes.<\/li>\n
                        • Business owners and founders<\/strong> can allocate resources with confidence, scaling what works and stopping what doesn\u2019t\u2014core to sustainable CRO<\/strong>.<\/li>\n
                        • Developers<\/strong> who understand experimentation concepts can implement cleaner targeting, reduce tracking bugs, and improve site performance during tests.<\/li>\n<\/ul>\n\n\n\n

                          Summary of Personalization Test<\/h2>\n\n\n\n

                          A Personalization Test<\/strong> is an experiment that measures whether tailored experiences outperform a control for specific audiences or contexts. It matters because personalization without proof can waste effort, harm trust, or mislead decision-making. In Conversion & Measurement<\/strong>, it provides causal evidence of incremental lift; in CRO<\/strong>, it\u2019s a repeatable method for improving conversion and customer value through relevance, not guesswork.<\/p>\n\n\n\n

                          Frequently Asked Questions (FAQ)<\/h2>\n\n\n\n

                          What is a Personalization Test in simple terms?<\/h3>\n\n\n\n

                          A Personalization Test<\/strong> compares a personalized experience against a standard one to determine whether the personalization causes better results (like more sign-ups, purchases, or qualified leads).<\/p>\n\n\n\n

                          How is a Personalization Test different from \u201cjust personalizing the site\u201d?<\/h3>\n\n\n\n

                          Personalizing the site changes the experience, but doesn\u2019t prove it helped. A Personalization Test<\/strong> uses a control group, randomization, and measurement so you can attribute changes to the personalization within Conversion & Measurement<\/strong>.<\/p>\n\n\n\n

                          What metrics should I use for CRO when running personalization experiments?<\/h3>\n\n\n\n

                          For CRO<\/strong>, pick one primary metric tied to value (conversion rate, revenue per visitor, qualified lead rate) and include guardrails like retention, churn, refunds, or lead-to-opportunity rate to avoid \u201cwinning\u201d the wrong way.<\/p>\n\n\n\n

                          How long should a Personalization Test run?<\/h3>\n\n\n\n

                          Long enough to reach a pre-planned sample size and cover normal variation (weekday\/weekend, campaign cycles). Duration depends on traffic volume and expected lift; narrow segments typically require more time.<\/p>\n\n\n\n

                          Can personalization hurt performance?<\/h3>\n\n\n\n

                          Yes. Bad targeting, inconsistent messaging, or overuse of incentives can reduce trust or margin. That\u2019s why guardrail metrics and solid Conversion & Measurement<\/strong> practices are essential.<\/p>\n\n\n\n

                          Do I need first-party data to run a Personalization Test?<\/h3>\n\n\n\n

                          Not always. Many tests can use contextual or on-site behavioral signals. However, stronger first-party data usually improves targeting accuracy and makes results more actionable for ongoing CRO<\/strong> work.<\/p>\n","protected":false},"excerpt":{"rendered":"

                          A **Personalization Test** is a structured experiment that evaluates whether tailoring an experience (message, content, offer, layout, timing, or journey) to a specific audience segment improves outcomes compared to a baseline. In **Conversion & Measurement**, it\u2019s not enough to \u201cpersonalize\u201d and hope for the best\u2014you need evidence that the change caused a meaningful lift, not just a nice-looking dashboard.<\/p>\n","protected":false},"author":10235,"featured_media":0,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[1889],"tags":[],"class_list":["post-7175","post","type-post","status-publish","format-standard","hentry","category-cro"],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/posts\/7175","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/users\/10235"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/comments?post=7175"}],"version-history":[{"count":0,"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/posts\/7175\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/media?parent=7175"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/categories?post=7175"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wizbrand.com\/tutorials\/wp-json\/wp\/v2\/tags?post=7175"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}