Activate It 2.0
Activate It helps you identify concepts with the highest activation potential — and learn how to make them even better. Use Activate It to assess potential for incremental growth, identify the right launch strategy and optimize your mid to late stage concepts for highest impact. Available for Fast Moving Consumer Goods, Quick Service Restaurants, Finance, Telco, and Tech.
Testing Process
- Monadic concept test to understand the growth potential of mid-late concepts
- Use a comprehensive set of KPIs to understand concept performance — including concept classification based on Trial and Breakthrough Potential. Tailored to industry specifics.
- Test up to 5* concepts with a sample size of 300 broad category users. This will unlock a broad and targeted sample to assess the overall opportunity (according to brand growth principles).
- Analysis can be done via side-by-side testing or using robust benchmarking analysis
- Includes deep dive into purchase behaviors, actionable attributes help identify opportunities for concept improvement to increase trial.
Set-up
Making Decisions with Activate It
- Key measures
- Utilizing a relevant benchmark
- Understanding the Concept Classification Framework
- Understanding your Concept Performance Indicators
- Understanding potential with Profiles
AI-Powered Concept Optimization
- What is it?
- Where to find it?
- What should I do with it?
- How does it work?
- What is a synthetic respondent?
- Why is it in Beta?
- What about security?
- Who owns the generated content?
AI Quick Reports
Other FAQ
Stimuli Guidelines
In Activate It, concept should have 4 key elements:
- A product / pack / feature image (photo or digitally rendered) which looks/feels like something consumers would see in-market.
- A brand name and logo, which are an important context that influences decision making.
- Bite-sized description, reflective of how people consume information and is more engaging than traditional long-form concepts. Focus on what’s important and describe it in a conversational manner. Include a benefit statement and emotional and rational justification. Description is not necessary for Tech, Telco and Finance products as it is usually part of the product visual.
- Product specifications. Information about flavors, variants, features, sizes/servings (including weight or volume), ingredients, where the it can be purchased, and price. Some of these specifications are optional inputs and vary by industry (CPG, QSR, Tech, Telco, Finance).
Price exposure is introduced later in the survey and it shouldn’t be included as part of the stimuli.
A full explanation of all stimuli and inputs can be found here.
Questionnaire Flow
Concept testing can look different from organization to organization, especially from industry to industry. We have three standard Activate It offerings for: 1) Fast moving consumer goods companies, 2) Restaurants and 3) Tech, Telco & Finance. The high level questionnaire flow is linked below.
View the Methodology Guide here.
Making Decisions with Activate It
Key Measures
- Trial (scaled Purchase Likelihood )
- Breakthrough (average of Distinctiveness & Advantage)
- Priced concept purchase likelihood
- Incremental growth evaluation
- Pricing & purchasing dynamics
- Concept diagnostics
Utilizing a relevant benchmark
Choose from three benchmarking approaches to assess your concept’s ability to outpace the competition.
- Normative benchmarks - category and country norms provide a broad view of how consumers rate your concepts in relation to a wide range of products within your category; a minimum of 20 concepts is required for a normative benchmark.
- User defined (tagged) benchmarks - a closely related group of concepts provide a precise view of how consumers rate your concepts in relation to close-in competitors and/or other products in your portfolio; a minimum of 5 concepts is recommended for a user defined benchmark.
- Direct benchmarks - when two or more concepts are tested together, statistical significance testing can be used to directly compare performance metrics between or across concepts.
Understanding the Concept Classification Framework
Strong potential with Early Adopters: Concepts listed as having "Strong Potential with Early Adopters" resonate with consumers that exhibit behaviors linked with early adoption. These concepts are listed by their Total Audience classification below. Early Adopters are consumers who try a new product or service when it first becomes available and in a shorter period of time than the average person. By understanding whether a new idea is appealing to Early Adopters, you can potentially identify breakthrough innovation, and no longer risk leaving strong ideas on the cutting room floor.
All concepts are classified into five "buckets," based on their performance across two key measures, Trial Potential and Breakthrough Potential:
Scale and sustain
These concepts have strong potential for scale amongst category consumers and should be prioritized. Develop full articulation and research to understand final commercial viability.
If concepts are Scale and sustain with strong potential with Early Adopters, they should be prioritized and exploited quickly before the opportunity to be a pioneer with this concept is lost.
Short term trial
These concepts have high competitive trial potential but aren't considered innovative enough in a category context to maintain excitement for long. Develop further and understand the short-term revenue opportunities where trial is needed but longer term potential is not essential.
If concepts are Short term trials with strong potential with Early Adopters, then Early Adopters see these concepts as being more breakthrough than category consumers. Develop further, considering how to highlight the breakthrough potential with a wider audience.
Seed and grow
These concepts have breakthrough potential but more modest trials. Consider moving forwards with slower/alternative approaches to growth such as different routes to market, limited distribution or a test-and-learn approach.
If concepts are Seed and grow with strong potential with Early Adopters, they have had their strong breakthrough potential confirmed by Early Adopters. Capitalize on the opportunity swiftly at a smaller scale through alternative routes to market, limited distribution or a test-and-learn approach.
Emergent
These concepts have a strong ability to drive distinction and advantage perceptions but low trial potential, suggesting a trend for breakthrough innovation. Consider monitoring and re-assess in the near future.
If concepts are Emergent with strong potential with Early Adopters, they have on-trend potential amongst Early Adopters, suggesting a trend for breakthrough innovation. Consider developing at a smaller scale to explore the opportunity.
Rework
These concepts have moderate-to-low breakthrough and trial potential amongst total audience category consumers. Either investigate potential amongst Audience Profiles, analyze, rework and retest, or deprioritize.
If concepts are Rework with strong potential with Early Adopters, then consider exploring further potential amongst Early Adopters to confirm trend and progress with a "seed and grow" perspective if confirmed.
More details on how the quadrant chart works here.
Understanding your key Concept Performance Indicators
Key Concept Performance Indicator chart shows "Top 2 Box" calculations
Each of these measures corresponds to a different question in the survey (summarized below). Use the configuration sidebar on the chart page to configure the scale measure display, profile/subgroup, and chart display type.
- Purchase Likelihood - unpriced concept: ...how likely would you be to buy [product] if it were available at a reasonable price? For QSR we ask about Order Likelihood, for Tech/Telco/Finance about Action Likelihood with an option to choose the desired action (use, buy, subscribe...)
- Purchase Likelihood - priced concept: The price for this product is [price]. Knowing this price, how likely would you be to buy [product]? We ask purchase likelihood for both unpriced and priced concepts to give a clean read on the power of the concept itself, and the impact of pricing.
- Distinctiveness: ...how similar or different is [product]? (assessed on a scale from very similar to very different)
- Advantage: ...agree or disagree with the following statement: [product/item/offering] is better than other products currently available
- The above are just the top three measures that feed the concept classification grid. We also report on all other relevant measures such as Relevance, Price Impact on Purchase Likelihood, Overall Emotion, Meets Needs, Clarity, Brand Believability, Substitutability, Category Believability & Commitment (FMCG), Brand Feeling (Tech, Telco, Finance)
Understanding potential with Profiles
You can use Profiles to define and create norms for your subgroups, then get a clear read on concept potential among them in your reports.
Powerful Diagnostics
Activate It offers a wide range of powerful diagnostics that help you optimize your concept:
- AI-powered concept optimization based on consumer feedback (for FMCG & QSR; Tech, Telco, Finance not available yet due to image editing limitations of LLMs)
- Auto-coded likes & dislikes for easy understanding of key themes
- Heatmaps of your concept image & text, for both likes and dislikes
- Emotional reaction to your concept, using a system 1 intuitive measurement with emojis as a universal language
- Attributes assessment
- Occasions (FMCG), Ingredients (QSR), Features (Tech, Telco, Finance)
- Assessment of concept variants
Assessment of Incremental Growth Potential
Activate It helps you ensure you drive incremental growth and minimize cannibalization of your portfolio with metrics like:
- Impact to Repertoire - helps you understand if respondents would buy the concept in addition to or to replace the brands they currently buy.
- Substitutability & Substitution Brands - a proxy to understand your sources of growth
- Potential to increase brand & category purchase frequency (for FMCG & QSR)
Purchasing & Pricing Dynamics
Insights that help you understand the different levers of growth - units, frequency, pricing, size/variant, ingredients (QSR), or features (Tech, Telco, Finance)
AI-Powered Concept Optimization
What is it?
The Activate It AI generated concept revision feature allows you to see what your concept description might look like if the feedback from your survey was applied directly onto the concept by an AI trained by Zappi.
The revised concept will also include an explanation from the AI of why it changed the concept in this way and visual recommendations outlining what elements to adjust on your packaging. You can generate AI revised concepts for the general category audience as well as for a specific target group (profile).
This is a new idea we are experimenting with and is available as a free feature in Beta.
Where to find it?
Activate It AI generated concept revisions are now available for Activate it 2.0 FMCG & QSR concepts. We are experimenting with making it available across a broader geographic portfolio but this is dependent on LLM's evolving capability to handle other languages.
You can download the AI concept revisions either by clicking on the banner in the top part of your concept analysis / reporting output
or download it straight from your concept tile (click on the three dots in bottom right of the tile and then ‘Generate revised variations’)
You're able to choose the target group for which you want to generate the optimized concept. If you wish to generate an AI optimization based on feedback from a specific target group, select a profile.
What should I do with it?
The purpose of the AI generated revisions is to give you a headstart on concept optimization. We encourage you to use it as a great starting point for revisions that helps you tweak the concept further more effectively based on your judgment.
We’ve seen significant improvements in performance of concepts optimized by our AI agent pointed at Zappi data. When testing the AI optimized concepts with consumers, 20% of key concept metrics saw significant improvement vs original concepts.
However, we know combining AI-driven concept optimization with human insight leads to even better results. Here's why:
- AI's Analytical Power: AI swiftly analyzes data to identify optimization opportunities.
- Efficiency: AI accelerates the optimization process, saving time and resources.
- Data-Driven Insights: AI uncovers patterns and trends for informed decision-making.
- Human Expertise: Humans provide contextual understanding and industry knowledge.
- Refinement and Adaptability: Humans apply creative thinking to refine concepts for maximum impact and adjust strategies based on changing needs.
- Quality Assurance: Human oversight ensures optimized concepts meet desired criteria.
So experiment with it, combine AI's analytical prowess with your expertise, and send us your feedback and suggestions so we can further improve the model. And don’t try it once, try it a few times, it will create something slightly new each time.
How does it work?
In order to revise the concept, we identify the dominant themes in consumer open ends (Likes & Dislikes) and analyze performance on selected KPIs - Clarity, Breakthrough and Brand Believability as these are the three areas where improvements are not easily identified from open ends. Then a LLM (Large Language Model) applies the results of this analysis to the revised concept, addressing deficiencies and amplifying the elements that were appreciated by consumers. These revisions are applied to the text description only (we anticipate making image adjustments in future versions).
We then instruct the LLM to evaluate the revision from the perspective of the target audience - the ‘synthetic respondent’.
What is the synthetic respondent?
For the synthetic respondent feedback, we are assigning the LLM a persona in line with the concepts target audience, and asking it for its impressions of the revision. In future we will have a full survey of synthetic respondents and we are also playing with hybrid models of synthetic and real respondents but again, this is our first iteration, tell us what you think!
Why is it in Beta?
Our new feature is a proof of concept for an AI concept optimization agent. In the future you will be able to push it in different directions but we have to start somewhere and this is our first feature, there are many more to come.
Large language models are groundbreaking, yet constantly evolving. Their occasional inaccuracies, or 'hallucinations,' are part of the learning curve—not just for the AI but for us as we integrate your invaluable feedback to enhance the feature's performance.
By participating in this beta, you're helping shape a smarter, more intuitive system. We're committed to transparency and improvement, working towards a future where our technology serves you with even greater precision and personalization. Thank you for joining us on this journey of innovation and growth!
What about security?
Your privacy and security are our top priorities. We adhere strictly to Microsoft Azure's comprehensive data security policy to ensure your information is safeguarded at all times. This means that any data you provide while using the feature is encrypted, securely stored, and never shared without your explicit consent. Your data (inputs), and your AI revised concepts (outputs) are NOT available neither to OpenAI nor to any third party.
Who owns the generated content?
When you use our feature to generate content, you have full ownership of the generated concepts. We're here to inspire and enable your creativity, not to claim it. Your success is our success!
AI Quick Reports
An AI Quick Report is an AI-generated report available upon the completion of a survey which provides a concise summary of results and diagnosis of why. It is produced by combining human expertise in how to interpret and diagnose consumer responses from your survey with AI.
AI Quick Reports are currently available for Activate It and Activate It Early on all industry variations for single concept analysis and multiple stims comparisons across one order. It will become available for comparisons cross order in the future. AI stories are written in English but can accommodate all countries of fieldwork.
FAQ for Activate It 2.0
A: Yes. We weight and normalize respondent data before putting it into the databases to ensure (a) data is consistent, (b) results are balanced and (c) data is discriminating to ensure an accurate representation of consumer behavior.
Q: Does Activate It screen out non-buyers in my category?
A: No. While the survey automatically populates screening questions based on the category you select and so is consistent each time you test in the same category, we do not screen out non-buyers in these questions. We instead weight key measures in reporting according to the sample's representation. This has two key benefits: 1) without screening quotas upfront, we’re able to field your surveys more quickly and cost-effectively, and 2) not screening out non-buyers allows you to analyze the non-buyer respondents and seek out white space opportunities or ways to influence this group specifically.
For example, the respondent will be asked a usage question like this one for the ready-to-heat meals category:
If they select "Never," they will continue with the survey like other respondents. However, the platform then automatically weights your data when the report is populated, based on the makeup of usage in the sample, so that your decisions are made based on the respondents you care most about. That way, if you were curious as to how your concept tested with those who selected "Never," you could always filter based on this criteria. Those with high performance among non-users of the category may represent an opportunity to attract new users to the category.