Many companies often find themselves in a quagmire of invalid leads when looking for leads, spending a lot of time and resources that are difficult to convert into real benefits. This article will delve into how to penetrate the fog during the PMF stage to accurately identify customers’ willingness to “pay” by constructing a double-layer funnel – “lightweight testing” and “low-cost value verification”.
PMF stage: How to penetrate the fog and identify the real “pay” willingness of customers?
Have you ever encountered such a dilemma?
- In the free or small POC stage, customers are enthusiastic and actively cooperate, and the result is goodAs soon as the official cooperation payment arrives, there is no follow-up, and even switched to inhouse to do it itself.
- clientFrequently ask for information and organize meetingsThe demand also seems reasonable, but it never gives a clear purchase answer. Time and energy are exhausted. A lot of manpower and material resources were invested in the early stage, and in the end only one sentence was exchanged“We’ll think about it again.”
- Seeing competitors win customers, but being “boiled frogs in warm water”, it was not until the end that they found out that the other party just regarded you as youFree consultation, for “Shop around” and then look for a supplier with a lower quotation for solution copying.
1. What is a value customer with payment potential?
If you are experiencing these, then the following content may help you clear the clouds, more accurately identify customers’ true willingness and ability to pay, and avoid the dilemma of “empty bamboo basket”.
In the product-market match (PMF) stage,Identify the customer’s genuine willingness and ability to pay, is far more important than getting the “ticket” of the POC. Because it is directly related to whether the resources you invest can be converted into real business value in the future.
So how to practice these “fiery eyes”?Here are some key screening methods and criteria:
- Clear needs and pain points: The business problems faced by our customers are highly aligned with our solutions, and they are motivated to seek change.
- resource investment capacity: Whether it is capital, time or manpower, customers are willing to invest certain resources to verify value.
- Decision-making influence: The customer has key decision-makers to support the POC and can promote the implementation in the future.
- Long-term cooperation potential: The customer’s industry or business is scalable, and there is a possibility of further deepening cooperation after the success of POC.
Specifically, the quantitative criteria:
1) The principle of strategic matching priority
- Industry position: The target customer needs to beSubdivision Top or core waist;
- Scenario typicality: The customer has a preliminary business problem to solve, and the demand scenario existsIt must be reusable(except for some industries such as finance);
- Technical adaptability: Existing product capabilities can be met60%The above core requirements
2) Hard indicator of payment ability
- Customer budget size: The annual budget associated with the demand for the solution≥ more than 3 times the estimated amount of the POC project
- Clarity of decision chain: It has been clearly dockedC-level or decision-making powerof the budget allocator
- Budget approval cycle: estimated1-3 monthsInternal budget approval process can be completed
3) Signal of willingness to cooperate
- Resource investment commitment: willing to provide business landing docking personnel and cooperate in providing the permissions, data or information required for POC, etc
- The time window is clear: can cooperate2-6 weeksValidation cycle
- Consensus on success criteria: Both parties have reached a written confirmation of the POC acceptance criteria
Typical counterexample customer portrait (elimination criteria)
1) Customer needs are vague X
Common trap descriptions: Customers cannot clearly describe their business pain points, or their needs are too broad to focus.
example“We want to improve overall operational efficiency”, “We need smarter solutions” (specific details are missing).
Avoidance strategy：
- Ask specific questions: Guide customers to quantify questions, such as “How much efficiency do you want to improve?” “How much is currently lost per month?” ”
- AIGC-assisted analysis: Analyze customer communication records, identify keywords and potential needs, and confirm whether there are clear goals.
2) Pretending to explore, but actually spying on military intelligence X
Common trap descriptions: Customers have no real purchasing intention, only to obtain competitive product information and understand industry trends.
example: Customers pay too much attention to technical details and competing product comparisons, and avoid talking about their own business needs.
Avoidance strategy：
- Emphasize unique value: Focus on the unique advantages of your own solutions and avoid falling into homogeneous competition.
- Value guidance: Shift the focus of the discussion to the customer’s business pain points and potential benefits, and guide the customer to share specific business scenarios.
3) Decision chain is missing X
Common trap descriptions: The level of communication objects is low, there is no decision-making power, or the customer’s internal decision-making chain is complex and lengthy, making it difficult to advance.
example“I will understand it first and then report upwards”, “We need approval from multiple departments internally”, and the decision-maker never shows up.
Avoidance strategy：
- Decision maker identification: Clarify the decision-making process and key decision-makers by asking questions such as “Who is usually responsible for these types of decisions?”.
- Strive for high-level participation: Take the initiative to communicate directly with the decision-making level, or invite senior management to participate in the POC plan discussion.
4) Unclear budget or insufficient budget X
Common trap descriptions: The customer does not have a clear budget account, or the budget amount is too low to cover the solution cost.
example“We don’t have a budget, let’s see the effect first”, “The budget is very limited, and we can only cover a small part of the cost”.
Avoidance strategy：
- Budget temptation: Ask directly about the budget scope and approval process, such as “What is the budget you expect for this project?”.
- Value disassembly: Adjust the solution according to the customer’s budget, provide a more cost-effective MVP version, and ensure that the customer can afford it.
5) The timeline is not clear X
Common trap descriptions: The customer does not have a clear plan for the start time and completion time of the project, lacks urgency, and is prone to delay.
example“Don’t worry, take a look before talking”, “Wait for the boss to come back and decide”, unable to give a clear time node.
Avoidance strategy：
- Time urgency guidance: Ask about the importance of the project to the client’s business, key time points, such as “How urgent is this project to your company’s business time node?”.
- Set a clear timeline: Work with the client to develop a POC timeline, clarify the goals and deliverables of each stage, and ensure that the project progresses as planned.
6) POC goal is not clear X
Common trap descriptions:P OC goals are too broad or disconnected from the actual needs of customers, making it difficult to evaluate POC results and fail to reflect value.
example“We want to improve overall performance”, “We need smarter solutions”, and it is not possible to give specific, quantifiable metrics.
Avoidance strategy：
- Goal focus: Define the core goals and key indicators of the POC with the customer, such as “What is the core goal of the POC?” “How to measure the success of POC? ”。
- quantifiable indicators: Transform goals into quantifiable indicators, such as improving efficiency by 20% and reducing costs by 15%, ensuring that POC results are evaluable.
7) Heavy dependency on existing scheme paths
Common trap descriptions: The customer sticks to existing processes and technologies and is reluctant to accept new solutions, even if the existing solutions are inefficient. Customers are overly reliant on existing processes, conservative about change, and difficult to accept innovative solutions.
example“We have always done this”, “This process cannot be changed”, “The risk of changing the plan is too great”.
Avoidance strategy：
- Dig deep pain points: Dig deeper into the limitations and potential risks of the customer’s existing solutions, such as data silos and poor scalability.
- Walk quickly in small steps: Customers are advised to start with small-scale pilots to gradually verify the value of new solutions and reduce transformation risks.
- Successful case showing: Share successful cases of similar customers in the same industry to eliminate customers’ concerns and enhance confidence.
8) Free Lunch Mindset X
Common trap descriptions: Customers only focus on free trials and POCs, and are unwilling to pay for solutions, hoping to gain value through long-term free use. Customers just want a free trial and are not willing to bear any cost.
example: Frequent requests for extended trial periods, added free features, and indefinite POC.
Avoidance strategy：
- Define the boundaries of payment: Clarify the scope, timing, and payment terms of the trial period and POC at the beginning.
- Value front: Provide core value during the trial period, allowing customers to fully experience the benefits of a paid plan.
- Timely charges: Stop the service immediately after the trial period ends, or charge a nominal fee to screen real customers.
2. How Deepseek assists in identifying customers with a real willingness to pay
The first level: lightweight testing – customer communication “looks, hears, asks”
Core: Through skillful questions and AIGC assistance, quickly identify customers’ real needs and willingness to pay.
The value and role of AIGC:
1) Pre-meeting preparation: AIGC-assisted problem design
- Based on the excellent question template of Top Sales, combined with the historical dialogue data with customers, AIGC assists in generating personalized question suggestions that are more suitable for business scenarios.Build an industry-customized question libraryto improve the pertinence and effectiveness of questions.
- By analyzing the question logic of high-success cases, AIGC will be in the questionEmbed budget association points.
- AIGC will simulate the customer’s decision-making chain and mark whether the problem can be reachedKey decision-makers’ concerns.
2) Post-meeting analysis: AIGC interprets the meeting minutes
- Give meeting recordings or minutes to AIGC and let it automatically analyze customer responses.
- AIGC can help you quickly identify customersPain point priorities, decision-makers, budget timeand other key information.
- AIGC can also help you determine whether the customer’s answer meets your settings“Qualifiedsignal”.
3) Follow-up: AIGC determines value customers and follow-up strategies
- Based on AIGC’s in-depth analysis of customer communication content,Automatically determine customer value quadrants (strategy, nurturing, relationship, observation)。
- AIGC combines industry knowledge base and success stories to serve customers in different quadrantsIntelligent recommendations for personalized follow-up strategies, such as “high-level relationship breakthrough plan”, “customized demand guidance content”, “competitor comparison analysis”, etc.
Example operation:
Step 1: Preparation before the meeting – AIGC-assisted question design
enclose“Where is the money, who has the final say, the bottom line of time”Three axes, design structuring problems.
The question isSpecific and quantifiable, avoid generalization.
For example: “What is your current approach to solving this problem?” How many resources do you spend each month? “What is the budget account corresponding to this demand?” “When will this problem have to be solved at the latest?” ”
Examples of AIGC prompts:
Please use the following information to generate a structured list of questions for initial communication with potential customers to quickly assess the client’s willingness to pay and the feasibility of the project.
Prerequisite information:
– Our goal is to have initial communication with potential customers to quickly assess their willingness to pay and project feasibility in order to decide whether to proceed with POC or more in-depth collaboration.
-The focus of this communication is to understand the client’s budget, decision chain and timeline.
– Problems need to be specific and quantifiable, and avoid generalization.
The question revolves around the following three core dimensions:
First Axe: Where is the money? (Budget)
Goal: Understand the customer’s budget source and scope, and determine whether they have enough budget to support the project.
Question template:
– “If we could help you solve [customer pain points], which budget pool would you plan to pay from?” Is it this year’s [department] budget or [project] special budget? ”
– “Has this budget been approved?” Or is there a flexible deployment space? How many are expected? ”
– “How does your company generally measure the return on investment (ROI) of similar projects?” ”
– “What is the approximate budget of the program? What specific business goals is this budget used to address this year? ”
– “If you need to pilot in a short period of time, how much can your budget cover? ”
Expected output (please generate specific questions based on the issue template):
-At least 3 questions related to budget sources.
-At least 3 questions related to the budget range.
Second axe: Who has the final say? (Decision Chain)
Target:Identify key decision-makers on the client side and understand their decision-making process.
Question template:
– “What process is generally required for this type of decision-making?” The last time you purchased a similar system, which leader was the last to approve the documents? ”
– “Which departments or personnel will participate in this budget approval and program evaluation besides you?” ”
– “Who is the final approver of this project?” Can you arrange for us to communicate directly with [Decision-Maker’s Name/Position]? ”
Expected output (please generate specific questions based on the issue template):
-At least 3 questions related to the decision-making process.
-At least 3 questions related to decision makers.
Third Axe: Timeline (Timeline)
Target: Understand the customer’s time plan, determine whether it is urgent, project start/go-live time.
Question template:
– “If our solution meets your needs, how soon can you expect to get budget approval or when can it be launched?” ”
– “How urgent is this project for your company’s business timeline? Is there a clear launch time? ”
– “When do you want to see initial results?” (e.g., POC results, scheme demonstrations, etc.)”
Expected output (please generate specific questions based on the issue template):
-At least 3 questions related to budget approval time.
-At least 3 questions related to the project’s go-live time.
Output Format Requirements:
– Please divide the generated questions according to “First Axe: Where is the Money?” “Second axe: Who has the final say? “The third axe: the bottom line of time? Organize three parts and mark the corresponding goals of each problem.
– Each question should have a clear direction and be easy to use directly in practical communication.
– For each question, please add examples of “pass signals” and “red flags” to quickly determine the potential value of the customer.
AIGC Sample Output:
First Axe: Where is the money? (Budget)
Target: Understand the source and scope of the customer’s budget to determine whether they have enough budget to support the project.
Issue: “If we could help you solve the problem of low sales conversion rates, which budget pool would you plan to pay from? Is it this year’s sales department operating budget or a special budget for digital transformation? ”
Qualified signals： Be able to clearly state budget items (e.g., “from the special fund for digital transformation”); There is a specific amount range (e.g., “within 200,000 yuan can be specially approved”)
Red flags: High uncertainty about the budget approval process (e.g., “Not sure, may need to request a new budget”)
Issue: “Has this budget been approved? Or is there a flexible deployment space? If it can help you increase your conversion rate, how much budget can you allocate? ”
Qualified Signals: The budget has been approved and has a clear budget scope (e.g., “approved, about 300,000 budget”)
Red flags: Not sure if approval is required, or the budget amount is ambiguous (e.g., “It may depend on the specific plan, the budget is not easy to say”)
Issue: “How does your company typically measure the return on investment (ROI) of similar projects? How much does increasing sales conversion rate account for your company’s KPIs? ”
Qualified Signals: Have a clear way to calculate ROI, and improving sales conversion rate is a key KPI (e.g., “We mainly look at the revenue growth brought by new customers, and conversion rate is the core KPI”)
Red flags: Unable to clearly describe how ROI is calculated, or sales conversion rate is not a key KPI (e.g., “This is difficult to measure, there are many KPIs”)
Issue: “What is the approximate budget of the program? What specific business goals is this budget used to improve this year? For example, customer unit price, such as user retention. ”
Qualified signals The budget of the program is clearly linked to specific business goals (e.g., “budget 500,000, goal is to increase the unit price by 15%”)
Red flags Budget and business goals are not strongly correlated (e.g., “I haven’t thought about this yet, let’s see the effect first”)
Issue: “If you need to pilot in a short period of time, how much can the budget cover? For example, how much budget can be supported for a one-month pilot? ”
Qualified Signals: The scope of the pilot budget can be clearly specified (e.g., “the pilot budget for one month is about 50,000”)
Red flags: Customers avoid talking about the pilot budget (e.g., “Can you do a free POC test to see the effect, and then talk about the budget”)
Second axe: Who has the final say? (Decision Chain)
Target: Identify key decision-makers on the client side and understand their decision-making process.
Issue: “What process does this kind of decision-making generally need to take? The last time you purchased a similar SaaS system, which leader was the last to approve the documents? What departments do you need to go through? ”
Qualified Signals: Be able to name the specific decision-maker’s name/position (e.g., “Sales VP Wang and Finance Director Li need to be double-signed”)
Red flags: “We will advance first, then report”
Issue: “In addition to you, which departments or personnel will participate in this budget approval and program evaluation? Will the technical department, or the marketing department, also be involved? ”
Qualified Signals: Be able to clearly identify relevant departments and people (e.g., “sales, marketing, and technical departments will participate in the review”)
Red flags: “Let’s talk about it then, let’s see the effect first”
Issue: “Who is the final approver of this project? Can you arrange for us to communicate directly with [Decision-Maker’s Name/Position]? ”
Qualified Signals: Take the initiative to arrange for you to communicate directly with decision-makers
Red flags: “The leader is busy, let’s discuss the details first”
Third Axe: Timeline (Timeline)
Target: Understand the customer’s time plan, determine whether it is urgent, project start/go-live time.
Issue: “If our solution meets your needs, how often can you expect to get budget approval or when can you get up and running at the earliest? How long does it take for general internal approval? ”
Qualified Signals: Give a clear time node (e.g., “must go live before the end of Q3”)
Red flags: “Don’t rush, do it first”
Issue: “How urgent is this project to your company’s business timeline? Is there a clear launch time? For example, must it be completed before Double Eleven? ”
Qualified Signals: Demonstrating time urgency (e.g., “must be resolved before the peak season next month”)
Red flags: “When the boss returns to China, I don’t know the time”
Issue: “When do you expect to see initial results? For example, an increase in conversion rates, or a change in user behavior. ”
Qualified Signals: Give a specific time expectation. (For example, “It would be great to see a 10% increase in conversion rates during the one-month trial period.”) ”)
Red flags: It is impossible to give a specific time expectation. (For example, “This depends on the situation, the sooner the better.”) ”）
Step 2: Post-meeting analysis – AIGC assists in the interpretation of meeting minutes
input: Communicate with customers about meeting recordings/minutes
AIGC prompts:
Analyze the Customer Meeting Conversation in depth to extract the following key signals:
1. Does the client mention specific problems, needs, and affordable costs for the current business? If so, extract the amount or time data involved.
2. Whether key roles in the decision-making chain are mentioned in the conversation (if so, list the departments/ranks mentioned)
3. Whether the customer gives a budget approval/delivery time window (yes/no, specific time)
4. Whether the customer has compared with competing products, or whether it has mentioned the centralized procurement process
Based on the customer meeting minutes, the sample output of the customer answer:
{
“budget_clarity”: 8,“budget_source”: “Digital Transformation Special Fund”,
“estimated_budget_range”: “30-500,000”,
“decision_chain”: {
“involved_roles”: [“IT Department Mr. Wang”, “Finance Department Mr. Li”],
“decision_maker”: “Report to CTO”
},
“urgency”: “high”,
“approval_timeframe”: “Approval is expected to be completed within 2 months”,
“deadline”: “must be online before Q3”,
“competitive_status”: “Not in contact with competitors”
}
Step 3: Follow-up – AIGC assists in customer value stratification and personalized follow-up strategies
After identifying and analyzing customer needs, we need to further stratify the value of customers to formulate accurate follow-up strategies. Next, we can combine the “four-quadrant customer classification method” to divide the value quadrant to which the customer belongs according to the results of communication and analysis with the customer.
AIGC Example Prompts:
Please further apply the following based on the above communication analysis with the customer [paste the conclusion of the meeting minutes of the second step]“Four-quadrant customer classification method”, which divides the value quadrant to which the customer belongs and the corresponding follow-up strategy:
AIGC Sample Output:
【Customer Type】Relationship customers, strategic customers
[Quadrant Reason]
– High budget clarity (7/10): Customers have a history of purchases, indicating budgetary support and willingness to pay, reducing sales risk.
– Low decision chain integrity (4/10): At present, only IT managers are in contact with it, which cannot effectively promote project decision-making, and there is a risk of missing the decision-making chain.
– High time urgency (9/10): It must be launched before Q4, and the time pressure is high, so it needs to be promoted quickly.
-Comprehensive judgment: Although the decision chain is incomplete, the customer has a clear budget and tight time, and the potential value is high, so it is judged as a relationship customer, but at the same time has the potential of a strategic customer.
【Action Suggestions】
1) High-level breakthroughs and establish direct communication:
Core objectives: Establish direct communication channels with the CTO or other higher-level decision-makers as soon as possible.
Specific measures:
– Ask the IT manager to make a referral and clearly express the need to communicate with senior management, emphasizing that it will help them better understand the project value and solution.
– Try to reach the target decision-maker directly through other channels such as internal company relationships and industry events.
– Prepare a condensed project presentation for senior decision-makers, highlighting strategic value and ROI.
2) Provide CEO endorsement cases to strengthen trust:
Core objectives: Leverage industry success stories to enhance the trust and recognition of solutions by senior decision-makers.
Specific measures:
-Prepare 2-3 representative success stories similar to the customer’s industry. The case content should include the project background, solution, implementation effect (preferably with quantitative data), and customer senior evaluation.
– When communicating with senior management, highlight these cases and highlight where they fit into the customer’s current needs.
– If conditions permit, strive to arrange for customers to communicate with the senior management of successful cases to further enhance persuasiveness.
3) Arrange decision-making level product demonstrations within two weeks to quickly advance:
Core objectives: Seize the time window and arrange product demonstrations for decision-making as soon as possible to demonstrate the core values and advantages of the solution.
Specific measures:
– Immediately confirm their needs and schedule for product demonstrations after communicating with senior management.
-Carefully prepare presentation content and focus on strategic value, ROI, risk control, etc. that decision-makers pay attention to.
– During the presentation, actively interact, answer questions, and offer trial or POC (proof-of-concept) opportunities.
4) Empower IT managers to drive:
Core objectives: While high-level breakthroughs are needed, IT managers are still important partners and need to be fully empowered to drive project progress.
Specific measures:
-Communicate regularly with the IT manager to understand the progress of the project and provide timely support.
– Share success stories, industry insights, and other information with IT managers to boost their confidence and recognition of the project.
– When communicating with senior management, mention the contributions of IT managers appropriately to help them establish prestige in the team.
5) Create a personalized follow-up plan:
Core objectives: Develop personalized follow-up plans based on customer specific circumstances to ensure efficient sales processes.
Specific measures:
– Clarify the goals, action steps, and responsible persons for each stage.
-Establish a regular communication mechanism to keep abreast of customer needs and feedback.
– Adjust sales strategies flexibly based on customer feedback to ensure that the deal is finally closed.
【Risk Warning】
-If effective communication with senior management is not successful, the project will face great resistance, and it is necessary to consider adjusting the strategy or re-evaluating the client as “observing the customer”.
-Time urgency is a double-edged sword, which can speed up decision-making and cause customers to abandon projects due to excessive pressure, requiring risk control.
Brief summary:
Through the above skillful question design (“where is the money”/”who has the final say”/”time bottom line”), as well as AIGC’s auxiliary analysis, Party B can quickly identify the customer’s budget source, decision-making chain and time pressure, and ensure that key information can be locked in the early stage of communication with the customer, thereby providing a basis for subsequent business negotiations and project promotion. Finally, through the four-quadrant classification method, you can see at a glance “whether to attack immediately or cultivate patiently”. At the heart of this combo is:With minimal communication costs, screen out “real demand customers” who are willing to pay for value, so that every follow-up action steps on the drum of the willingness to pay.
Layer 2: MVV low-cost value verification design
MVV (Minimum Value Verification) It is the key sieve for “identifying value POC customers” – after the initial judgment of “looking, hearing, asking”, it uses the lowest cost to help you see clearly: whether the customer is “just looking at it” or “really wants to pay for it”. If the customer takes the initiative to cooperate with the verification and ask about the effect, it isA beacon of real willingness to pay。 At its core, it is:Use a quantifiable “small value goal” to prove that “your solution solves my specific problem”, like using a lighter to burn the fabric, you can immediately see whether it is really flame retardant. Let you enter the POC before you enter the POCTarget high-potential customers who are willing to take action for value。
MVV (Minimum Value Verification)
definition: Show solutions to customers at the lowest costquantifiable business valueFor example, give AI models a 3-day trial or open a core function module.
Core objectives: Let customers go from “hearing that it is useful” to “seeing the effect”, and simplify the complex decision of “whether to buy or not” into a single point judgment of “whether this function is worth X yuan”.Reduce pre-sales explanation costs by 80%。
Applicable scenarios: It is especially suitable for algorithmic products (such as AI prediction models) or visualization tools (such as data dashboards), such as demonstrating to retail customers “can AI slow-selling warning models reduce end-of-quarter losses by 20%”, which is more impactful than a 20-page PPT.
Key elements required for MVV design:
- Validate the target: Clarify the purpose of the validation, the problems and the acceptance criteria for success or failure.
- Core indicators: A quantifiable metric used to measure the success of the validation.
- Validation cycle: Set a time frame for verification, usually within 3 weeks.
- Resources required: The data, permissions, or information that the customer must provide.
- deliverables: The final deliverable, which typically includes validation reports and data analysis tools.
example: Provide MVV: “Can the AI slow-moving early warning model reduce the liquidation loss at the end of the quarter by 20%”.
How can AIGC assist in output and implement MVV in the shortest time:
- Intelligent MVV Generation:Automatically design verification paths based on customer needs
- Automated Data Insights:Parse customer-provided data to generate visual conclusions
- Dynamically adjust validation metrics:Based on the initial results, the subsequent verification direction is optimized
Design MVV scheme
AIGC Prompt Template:
MVV design template
Verification objective: Prove that the solution can solve [specific problem]
Core indicators: [single quantifiable indicator] (e.g., 20% reduction in manual reconciliation time)
Verification period: ≤2 weeks
Resources required:
– Client is required to provide: [up to 3 required data/permissions]
– Our investment: [Clarify manpower/time limits]
Deliverable: 1-page verification report (including before and after screenshots)
MVV solution example: Identify slow-moving funds in advance and optimize the quarter-end clearance strategy
Step 1: MVV design scheme output:
AIGC prompts:
Customer pain points: Loss of profits due to slow sales at the end of the quarter
Industry: Footwear e-commerce
Data availability: 50 SKU sales data (daily sales, inventory, discount rate data) can be provided.
Please generate an MVV scenario:
Refer to the MVV design template:
Verification objective: Prove that the solution can solve [specific problem]
Core indicators: [single quantifiable indicator] (e.g., 20% reduction in manual reconciliation time)
Verification period: ≤2 weeks
Resources required:
– Client is required to provide: [up to 3 required data/permissions]
– Our investment: [Clarify manpower/time limits]
Deliverables: [1-page verification report (including before and after screenshots)]
AIGC Outputs:
Validate the target
Verify that the AI slow-moving early warning model can identify 80% of the slow-moving funds 15 days in advance to reduce the loss of liquidation at the end of the quarter.
Core indicators
Early warning accuracy: Number of successful warnings / Actual unsalable amounts ≥ 80%
Value quantification: The increase in the discount rate of successful early warnings × Quarter-end inventory
Validation cycle
≤ 2 weeks (from data collection to model deployment)
Resources required
Provided by the customer：
– Daily sales, inventory, discount rate data for 50 SKUs for the quarter (Excel format)
-The commodity supervisor marks the actual unsalable amount every day
We are investing：
-Data cleaning and preparation (1 person-day)
-Model training and optimization (GPU automatic scheduling, zero labor)
deliverables
Unsalable warning signboard：
-Dashboard showing success rate and warning effect in real time
ROI Calculator：
– Customers can enter the discount rate to automatically calculate the estimated revenue
Step 2: MVV deliverables output
Practical steps:
1. Data preparation and cleaning
target: Ensure data quality to facilitate AIGC to perform analysis and model training.
Steps:
1) Input data format:Customers are required to provide an Excel file containing daily sales, inventory, and discount rate data for 50 SKUs for the quarter.
2) Data cleaning:
- Missing value handling:Fill in missing values or delete data anomalies (such as missing values in sales data or illogical inventory data).
- Outlier detection:Utilize AIGC to automatically identify and flag abnormal data (such as sudden stock outages or sales spikes).
AIGC prompts：
AIGC prompts: data cleaning and outlier handling
“””
Please process the sales_data.xlsx provided by the customer, requiring:
1. Delete the null line or fill in the null value.
2. Mark SKUs with abnormal sales volume, inventory, and discount rates (if inventory is negative and sales volume is greater than the actual maximum).
3. Output the cleaned data and generate a cleaning report.
“””
AIGC output：
Cleaned data table：
(Excel or CSV file, mark the results of outlier or missing value processing)
Cleaning report：
Data Cleaning Report:
– Removed 3 missing value rows to fill in the gaps in the sales data (method: fill with average).
– Marked SKU 103 inventory as negative and requires further verification.
– The cleaned data has a total of 50 valid SKUs ready for slow-selling analysis.
2. Slow-moving goods prediction and model training
target: Use AI models to predict which styles are likely to be slow-moving and optimize prediction accuracy.
steps：
- Unsalable probability calculation: Based on historical sales data, AIGC passedSales decay curveCalculate the unsalable probability of each SKU and identify the styles with a higher risk of slow salage.
- Model training: Train historical data using deep learning (e.g., LSTM neural networks) or machine learning (e.g., random forests, XGBoost) models to predict future slow-moving payments.
The role of AIGC in this link
- The code is automatically generated: Based on the prompt, AIGC generates a complete code template for data preprocessing, feature construction, model training, and prediction.
- Process guidanceAIGC provides detailed operation instructions for each step, helping to understand the entire process from data cleaning to model evaluation.
- Output documentation: The model evaluation report and prediction results generated by AIGC simulation provide a basis for subsequent business verification and report writing.
Examples of AIGC prompts：
# AIGC prompt: slow-moving prediction and model training
“””
Based on the sales data file sales_data.xlsx provided by the customer, complete the following tasks:
1. Calculate the unsalable probability of each SKU based on historical sales data. Use the sales decay curve model to analyze the recent sales of each SKU and output the predicted slow-moving probability.
2. Use a machine learning model (e.g., XGBoost) to train a classifier that predicts which SKUs are likely to be unsalable in the future. Set the warning threshold to 80% (that is, SKUs with a predicted slowdown probability of 80% or more than the model is considered as an early warning payment).
3. Generate the following output:
– A list of slow-selling forecasts showing the probability of slow-selling for each SKU (e.g., SKU 101: 85%, SKU 103: 92%, SKU 104: 70%).
– Model evaluation report with accuracy, precision, recall and F1 Score, indicating the model used and its parameters.
Please output the complete Python code and the simulation run results for subsequent reference.
“””
AIGC output：
Generated Python code
Here’s a sample code that simulates the entire process from data loading, slow-moving probability calculations, model training, to generating prediction results and evaluation reports. You can run this code directly in a Jupyter notebook or other Python environment.
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
import xgboost as xgb
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
import matplotlib.pyplot as plt
# 1. Data loading
# Assume that the sales_data.xlsx contains fields: SKU, sales, inventory, discount
df = pd.read_excel(“sales_data.xlsx”)
# 2. Data cleaning
df.dropna(inplace=True)
df = df[df[‘inventory’] >= 0] # Delete records with negative inventory
# 3. Unsalable probability calculation
# Assuming a sales decay curve: compare recent sales with historical average sales,
# A simple example here: If the sales volume is below 80% of the historical average, the probability of slowing down is calculated: 1 – (sales / mean_sales)
mean_sales = df[‘sales’].mean()
df[‘Unsalable Probability_Sales Decay’] = 1 – (df[‘sales’] / mean_sales)
# Limit the probability range between 0 and 1
df[‘Unsalable Probability_Sales Decay’] = df[‘Unsalable Probability_Sales Decay’].clip(0, 1)
# 4. Construct slow-moving labels
# Slow-selling volume below the median: Sales below the median is marked as 1 (unsalable), otherwise it is 0
median_sales = df[‘sales’].median()
df[‘is_unsold’] = (df[‘sales’] < median_sales).astype(int)
# 5. Feature selection and data division
features = [‘sales’, ‘inventory’, ‘discount’, ‘unsalable probability_sales decay’]
target = ‘is_unsold’
X = df[features]
y = df[target]
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
# 6. Model training: Use the XGBoost classifier
model = xgb.XGBClassifier(use_label_encoder=False, eval_metric=’logloss’)
model.fit(X_train, y_train)
# 7. Forecasting and evaluation
y_pred = model.predict(X_test)
accuracy = accuracy_score(y_test, y_pred)
precision = precision_score(y_test, y_pred)
recall = recall_score(y_test, y_pred)
f1 = f1_score(y_test, y_pred)
print(“Model Evaluation Report:”)
print(“Accuracy:”, round(accuracy, 4))
print(“Precision:”, round(Precision, 4))
print(“Recall:”, round(recall, 4))
print(“F1 Score：”, round(f1, 4))
# 8. Outputs the slow sales forecast probability for each SKU
# Obtain the prediction probability of the test set SKU, assuming that the probability predicted by the model is 1, which is the slow-moving probability
pred_probs = model.predict_proba(X_test)[:, 1]
results = pd.DataFrame({
‘SKU’: df.loc[X_test.index, ‘SKU’] if ‘SKU’ in df.columns else X_test.index,
‘Unsalable Probability’: np.round(pred_probs * 100, 2) # Converts to a percentage expression
})
print(“\nSlow-Selling Forecast Results (Partial):”
print(results.head())
# 9. Visual display (optional)
plt.figure(figsize=(8, 4))
plt.hist(pred_probs, bins=20, color=’skyblue’, edgecolor=’black’)
plt.title(“Unsalable Probability Distribution”)
plt.xlabel(“Unsalable Probability”)
plt.ylabel(“SKU Quantity”)
plt.show()
Slow-moving sales forecast results：
Slow-moving forecast list:
SKU 101: Probability of slow salable = 85%
SKU 103: Unsalable Probability = 92% * (Negative inventory marked as abnormal)*
SKU 104: Unsalable probability = 70%
Model evaluation report:
Model accuracy: 87.5%
Model used: XGBoost (classifier)
Accuracy: 0.88
Recall rate: 0.83
F1 Score：0.85
Warning Threshold: 80%
3. Data insights and visualization
target: Generate data analysis reports and visual charts that can help customers make decisions, providing intuitive business insights.
steps：
1) Identify slow-moving money: Automatically identified by AIGCUnsalable payments, lists the TOP 10 slow-moving styles to help customers clearly understand which styles need to be dealt with in advance.
2) Generate visual charts：
- Inventory vs. sales chart: Shows the trend of inventory and sales of slow-moving goods.
- Discount rate distribution chart: Displays the change of discount rates for different styles to help customers understand the impact of different discount rates on slow-moving items.
Examples of AIGC prompts：
# AIGC prompt: Generate data insights and visualization charts
“””
Generate the following visualizations based on customer data:
1. Comparison chart of inventory and sales of the top 10 styles with unsalable probability.
2. Distribution chart of the year-end clearance discount rate of different styles.
3. Identify the correlation between inventory volume, sales volume and discount rate of slow-moving goods.
“””
AIGC output：
Comparison chart of slow-moving inventory and sales(Chart format):
A bar chart showing the inventory and sales of each slow-moving item is as follows:
SKU 101: Stock 20, Volume 500
SKU 102: Stock 15, Volume 600
Distribution chart of the discount rate at the end of the quarter(line chart or bar chart):
Displays the distribution of discount rates by style, highlighting slow-moving payments with higher discount rates.
4. ROI Calculator Design and Delivery
target: Help customers quantify the business value of their optimization schemes through ROI calculators, especially in terms of discount rate adjustments.
The role of AIGC:
- Excel template: An interactive ROI calculator can be delivered quickly, allowing customers to enter relevant parameters and automatically calculate recovery losses.
- Online calculation tools: If customers need web tools, AIGC can help generate online computing tools based on HTML and JavaScript to achieve real-time calculations.
AIGC Outputs:
a. Generated Excel file: ROI calculator template
Design ideas：
Set up the input box in Excel to allow the customer to enter the following parameters:
- Unsalable inventory: Customers can enter the unsalable inventory quantity (e.g. 3200 pairs).
- Original discount rate: Customers can enter the original discount rate (e.g. 40%).
- Optimized discount rate: Customers can enter an optimized discount rate (e.g. 60%).
- Customer unit price: Enter the price of each pair of shoes (e.g. 300 yuan).
Formula settings：
Calculation of recovery loss: Calculate the recovery loss by Excel formula, the formula is:
Recovery Loss = Slow Inventory × (Optimized Discount Rate – Original Discount Rate) × Unit Price
According to the above formula, the Excel file can be calculated automaticallyRecover losses。
Excel template content: Suppose the Excel file is designed as follows:
formula：
挽回损失 = 3200 × (0.6 – 0.4) × 300 = 19.2万元
Screenshot of Excel file(Simulation example):
Excel formula implementation: Enter the following formula in the Recover Loss cell:
= B3 * (B4 – B2) * B5
- B3: Unsalable inventory
- B2: Original discount rate
- B4: Optimized discount rate
- B5: Customer unit price
b. Online calculation tools (optional)
AIGC can also help with generationOnline tools(such as JavaScript-based web tools) to achieve ROI calculation.
Generated online tools (HTML/JavaScript examples)：
Here’s a simple HTML and JavaScript code to generate the ROI calculator:
<!DOCTYPE html>
<html lang=”en”>
<head>
<meta charset=”UTF-8″>
<meta name=”viewport” content=”width=device-width, initial-scale=1.0″>
<title>ROI Calculator</title>
</head>
<body>
<h1>Slow-moving inventory recovery loss of calculator</h1>
<label for=”discount_original”> current discount rate:</label>
<input type=”number” id=”discount_original” value=”40″ min=”0″ max=”100″>%
<br>
<label for=”discount_optimized”> Optimized discount rate:</label>
<input type=”number” id=”discount_optimized” value=”60″ min=”0″ max=”100″>%
<br>
<label for=”inventory”> slow-moving inventory:</label>
<input type=”number” id=”inventory” value=”3200″>
<br>
<label for=”price”> Unit price of the customer:</label>
<input type=”number” id=”price” value=”300″>
<br>
<button onclick=”calculateROI()” > calculate the recovery loss</button>
<h2>Recovery of losses: <span id=”result”>0</span> yuan</h2>
<script>
function calculateROI() {
var discount_original = document.getElementById(“discount_original”).value / 100;
var discount_optimized = document.getElementById(“discount_optimized”).value / 100;
var inventory = document.getElementById(“inventory”).value;
var price = document.getElementById(“price”).value;
var recovery_loss = inventory * (discount_optimized – discount_original) * price;
document.getElementById(“result”).textContent = recovery_loss.toFixed(2);
}
</script>
</body>
</html>
Function description:：
- User inputCurrent discount rate, optimized discount rate, slow-moving inventoryandCustomer unit price。
- After pressing the “Calculate Recovery Loss” button, the page displays the calculated recovery loss amount.
Both forms of tools help customers clearly see the financial impact of optimizing discount rates on slow-moving payments when making decisions and make more precise business adjustments.
5. Customer report generation
target: Provide clients with a concise and clear report outlining MVV validation results and value quantification.
steps：
Report content：
- The actual number of slow-moving amounts and the accurate number of early warnings: Provide the actual situation and early warning accuracy of slow-moving funds.
- The impact of increased discount rates: Shows the effect of the discount rate.
- Increase revenue estimates: Show the potential increase in revenue through the ROI calculator.
Examples of AIGC prompts：
AIGC prompt: Generate customer reports
Based on the following validation results, generate a customer report:
– Actual slow-moving amounts: 32 models
– Accurate number of early warnings: 28 models
– Early warning accuracy: 87.5%
– Average discount rate increase: from 4% to 60%
Please make:
1. 1 page PPT summary (with comparison charts and key data)
2. Loss Calculator (allows customers to adjust parameters)
“””
AIGC output：
Customer Report (PPT Summary)：
Slide 1: Summary of verification objectives and results
– Identification of 87.5% of slow-moving payments 12 days in advance
-Quarterly revenue increase: 192,000 yuan
Slide 2: Comparison chart (changes in inventory, sales volume, and discount rate of slow-moving goods before and after optimization)
Loss Calculator：
Parameters such as discount rate, inventory volume, and customer unit price can be adjusted to calculate and recover losses in real time.
Deliverables examples：
【Verification Conclusion】
✓ Identification of 87.5% of slow-moving payments 12 days in advance
✓ Estimated Recovery Loss:
Current season inventory: 3200 pairs
× Average discount increase: 20%
× Unit price: 300 yuan
= Quarterly revenue increase: 192,000 yuan
Step 3: Behavioral signal scorecard to continuously track MVV feedback
Testing Standards:
In the process of identifying value POC customers,The Behavior Signal Scorecard is like an accurate monitor, continuously track MVV feedback, so that the customer’s true intention is clearly visible in the dynamic score.
| Behavioral Signals | Score | Detection Methods |
| Proactively ask for implementation details | +2 | There are 3 related questions ≥ in the meeting minutes
| Timeout Provide Data | -1 | Delay days× 0.5 (max -3) |
| Invite other departments to participate in the meeting | +3 | The number of new departments × 1 point |
| Request Adjustment Validation Metrics | +2 | Revised to be stricter |
| The total score of ≥ 5 points continued to advance, and the ≤ 2 points were terminated
Through the above steps, AIGC can not only efficiently complete data analysis, model training, visual display and report generation, but also dynamically optimize verification indicators and adjust the verification direction in real time. This method greatly improves the implementation efficiency and accuracy of MVV, helping customers quickly verify the commercial value of the solution while reducing the cost and cycle of verification.
Summary:
In the journey of finding value customers in the PMF stage, we need to clear the fog of “opportunism” and directly hit the willingness to pay “real money”. By building a double-layer funnel: with “lightweight exploration” customer communication as the initial screening, AIGC empowers question design and minutes analysis to quickly capture key signals such as budget and decision-making chain; Then use “MVV low-cost value verification” as a fine filter to test the customer’s determination to pay for the solution with the least investment. This combination is designed to avoid ineffective investment and accurately target “action buyers”. The core intention of this methodology is:Refuse to “cast a wide net” of ineffective investment, so that enterprise resources can accurately flow to high-value customers who are “willing to vote with their budgets”, transform the sales funnel into a “value filter” Grasp the willingness and ability of customers to pay in order to drive growth and come out on top in the fierce market competition.
