Automating Customer Feedback Analysis with Google Forms, AI, Google Sheets, and WhatsApp

The first step in automating customer feedback analysis is collecting feedback from your customers. You will use Google Forms to help you collect all feedback in one place without digging through all your emails, messages, and chats. With Google Forms, you can easily collect and organize customer ratings, comments, and contact information in Google Sheets without any extra effort from you or your team.

The form that you build in Google Forms is the initial part of this process and establishes how the process works overall. The way to build this form correctly is as follows:

1. Go to Google Forms, create a new form, and select the "Customer Satisfaction Survey" from the available template options or create a new form from scratch.
2. Add the necessary fields to your form: A customer rating from one to ten (using the Linear Scale question type), space for detailed comments (using the Paragraph question type), and customer contact information (email with validation).
3. Connect your Google Form to a Google Sheet so that the responses automatically save there.
4. Set up a branching logic system so that if the customer rated you a four or lower, they will be asked a follow-up question regarding the reason for their rating. If they rated you an eight or higher, they will be asked to provide a detailed testimonial that may be published on your website or social media channels.
5. Create a confirmation message that will be sent after the customer submits their survey to let the customer know when they can expect to hear back from you.

All of these small details can be very impactful.

For example, one crypto company followed up after their token was listed with a three-question follow-up survey asking: "Please rate the ease of trading the token," "What do you think could be improved in our documentation?" and "Would you recommend the project to others?" Within one week, 1,200 surveys were submitted. The form was clever because it highlighted people with ratings under six and gathered 340 specific comments instead of just saying, "I didn't like it"—so it's a positive experience for users. Users of a SaaS application provided feedback through the form after every project, and 78% of them specified on which project they used the product to solve their task. That type of information was very valuable for writing case studies and creating training materials. Additionally, they had an invisible field for collecting the UTM tags to see which channels produced the most committed customers.

Tip: Don't ask too many questions. According to user research, each additional mandatory question decreases the response rate by 5% to 8%. If you ask only three questions, the response rate improves by about 7%. Having a progress bar shows users what they have left and encourages users to complete all fields. Small messaging such as "Enter your email to receive an answer within 24 hours" increases response rate by 12% to 15%.

When creating surveys, try various question types. For example, using checkboxes is a simpler approach to reduce cognitive load and make it easier to analyze results. Timing is critical, so you should deploy surveys following moments of peak customer engagement (i.e., following purchase, course completion or resolution of a customer issue). In these instances, your response rate could be three to four times higher than under other circumstances.

It's exciting to understand how to use artificial intelligence (AI) to analyze customer feedback data. For example, it takes just seconds for AIs to break down customer feedback similar to this: it categorizes sentiments (positive, negative or neutral) and ranks the main themes (e.g. price, quality, or delivery) along with customer feedback types (e.g. complaint, suggestion, and praise). When using AI, it takes a person several minutes to perform the same task for a single customer feedback item. The speed at which processes are completed and their completion rates are significantly increased using AI. This is due to it being approximately 90 times faster than traditional processing and the fact that it removes the chance for human error and fatigue.

Some obvious benefits would be:

• An AI can review a thousand reviews in less than half a minute, meaning that three analysts would take over 20 hours to accomplish the same task.
• The rules associated with using AI are consistent for everyone, therefore, there is no bias whatsoever.
• In the process of using AI, it may discover subtle connections between items such as the negative feedback concerning delivery related to a region and/or a certain time of the day.
• Scalability is another benefit of using AI; if you set it up to process 10 reviews, you'll be able to easily increase that to 100,000 reviews without losing quality.
• AI is capable of working in over 50 languages, making it important for any type of global project.
• AI is a budget saver; instead of hiring a team of analysts to review each review, you are able to invest in actually improving your product.
• The current state of Natural Language Processing (NLP) has progressed to the point of being able to do a tremendous amount – from simply classifying to determining the context of a sentence, and even catching sarcasm.
• In addition, sentiment analysis captures the subtleties of emotion, like when someone comments, "Finally got my problem fixed!" The system would tag that comment as having positive sentiment associated with relief.
• AI can also recognize and capture entities such as names, dates, and brands using Named Entity Recognition and therefore link feedback to events associated with a product or service.
• With Topic Modeling, you can use AI to automatically group reviews based on themes, instead of manually labeling each review.
• Keyword extraction allows the support team to quickly identify repetitive issues or concerns for immediate attention.
• Intent Classification identifies what the action should be taken based on the review: resolve complaints, expedite resolutions, or add to the backlog and address the suggestion in the future.

If configured properly, according to 2023 Gartner research, the accuracy of AI related to sentiment detection can be as much as 94%-98%. The practical application of these benefits is also impressive. In the case of the project arbitragescanner.io, Artificial Intelligence indicated roughly a quarter of the negative reviews were due to delays on updating trading price on exchanges. Once the issue had been addressed the share of negative feedback dropped from 23% to 4%. AI subsequently recommended an integration with Telegram resulting in an increase in revenue of 18%. Similarly, a Web3 marketing company analysed 40% of its large-scale customers complaining of being unable to access real-time analytics. As a result, this company developed and implemented a live dashboard resulting in a retention rate increasing from 62% to 81% and an increase in the average revenue per customer of $7,000.

Integrating And Storing Data Within Google Sheets

Google Sheets is not simply a spreadsheet; it is a central repository for keeping track of all your forms, results generated via Artificial Intelligence, workflow statuses, etc. As a result, it becomes an ongoing living dashboard that allows every team member to view daily trends, favourite issues, and the status of all company activity without having to make additional phone calls or send endless emails.

• Have you already connected Google Sheets with Google Forms?
• Have you already established a method for transferring data from Google Sheets to Artificial Intelligence? A variety of tools exist for this purpose, such as Zapier, Make.com, and Apps Scripts.
• Create three additional columns for Artificial Intelligence results: sentiment, theme and urgency.
• Create a column titled "Processing Status" to allow for quick identification of items that require immediate attention.
• Integrate both WhatsApp and Telegram via API, so that managers are immediately notified of any negative review.
• Utilize conditional formatting on Google Sheets to colour-code the sentiment of each review for quick reference and easy identification.

Use Pivot Tables or Charts in Google Sheets to allow for quick assessment of historical performance. For example, if you notice a gradual decline in your company's average rating over a three-month period, it could be a sign that action is required. Moreover, histograms can be useful in identifying customer groups who may potentially be ready to churn. Calculating various metrics (such as NPS) in spreadsheets can easily be done using a COUNTIF formula, which provides great insight into customer loyalty. An important benefit of utilizing Apps Script to automate this process is the ability to send notifications; for example, when a review with less than 5 stars is entered into the table, the manager gets a notification. This drastically reduces the response time from hours down to 20 minutes.

As soon as a client provides feedback, it is critical to immediately follow up with a thank you or an apology via WhatsApp or email. The message should explain what will happen next. This entire process can be automated without sacrificing the personal touch.

1. In order to connect with WhatsApp Business API through Meta Business Suite, create a business account.
2. Next, verify your phone number and obtain a token that will allow you to send messages.
3. Set up a webhook for receiving incoming messages and automatically responding.
4. Prepare message templates and submit to moderation. If moderation is not passed, messages will not be sent.
5. Bind the WhatsApp API with Google Sheets so that managers are notified of negative reviews immediately.
6. To determine next steps based on the rating received: if the client leaves a negative review, the manager will call; if the client leaves a positive review, the client will receive a thank you and a link for referral purposes.

For emails, we find that the best time to send a feedback request is approximately 1-2 days after the service was received when the client still remembers the details about their experience. If possible, personalize the subject line of the email. Rather than simply writing "Leave a review," utilize a subject line such as "Ivan, how did we help you?" This can improve the open rates by 20-25%. When including the link to the form, include it in the first paragraph. The more clicks required to get to the form, the more likely clients will abandon the process (approximately a 10-15% drop in conversion). To encourage people to submit reviews, add social proof to the review form stating: "1,200+ clients have submitted their opinions already!" Place a time estimate of a couple of minutes on the review form so clients can clearly understand how much time they need to devote to it. As you experiment with your layout, the results will vary based on the type of business you operate.

Some examples of how the automated system allows the process to be automated are:

1. Positive Review: A client gives a rating of 9 and says; "Great Service!" The AI identifies the positive sentiment, and sends a 'thank you' message via WhatsApp, along with a link to the referral program. The manager sees an opportunity to upsell the client.
2. Negative Review: A client gives a rating of 3, due to difficulties with delivery. The AI recognizes the negative sentiment and automatically sends an email with an apology and a promise to contact the client within two hours. The manager receives a detailed alert via Telegram.
3. Neutral Review with Suggestion: A client submits a neutral review, giving a rating of 7, and makes a suggestion for a new API feature. The AI recognizes this as a suggestion and sends a 'thank you' email with a link to the update log. The manager gets notified and the request goes into a special priority table.

Key Benefits of Automating Feedback Analysis

Speed up processing of Reviews: When a client manually processes 500 reviews, the average processing time is approximately 20 hours. Using the automated process, that same 500 reviews can now be processed in 5 to 10 minutes or 90 to 100 times faster than manually processing. The client receives a response from the company approximately 5 minutes after he/she completes the review, and this has been shown to improve the NPS score from 8 to 12 points. Reviews that received a score of less than 5 are given priority by the system. In the case of one cryptocurrency project, this improved the retention of 12% of the clients who had indicated they would be leaving due to a lack of response from support. A quick feedback channel reduces client churn by 15% to 20% and improves a client's loyalty to your company.

Enhanced Customer Care: Artificial intelligence provides a way to find problems that often get overlooked by contact centre agents, such as 18% of negative feedback relating to a confusing interface, which can often be expressed in several different ways (for example: "I didn't understand," "I was lost"). The AI will group all of the negative comments together by theme to allow your team to address the issue with the user experience. By leveraging the historical interaction data automatically, your bot can provide personalized responses. For instance, instead of saying "Thank you" in a generic manner, the bot might say, "Andrew, we are glad your API issue is resolved! If there is anything else we can help with, please let us know."

Lower Cost of Data Processing: Analyst salaries can range from $2,000 to $4,000 a month, while the average cost of automation ranges from about $50 to $200 per month and saves the company almost $3,800 while producing the same quality of data. Report preparation time has gone from 8 hours to 10 minutes using pre-programmed dashboards. According to an independent study, 1,000 manually processed reviews are 3-7% inaccurate due to human error, while AI is 94-98% accurate. This is particularly relevant in regulated industries where mistakes may have serious consequences.

Core Use Cases & Examples of Automation

E-commerce: Customers receive an email with a feedback form after each order. AI identifies the primary concerns: packaging, product condition, courier, etc. As a result of this, one company changed its regional delivery partner and decreased negative feedback from 22% to 6% and increased the likelihood of repeat purchases by 19%.

SaaS: Organisations typically send surveys after concluding a free trial. AI is able to analyse customer reviews for issues such as missing features or having a challenging interface. By listening to its customers and adding requested features, one company increased the percentage of customers converting from free trials to paid subscriptions from 18% to 27%.

Cryptocurrency Projects: Individuals who utilize arbitragescanner.io provide feedback to one another on a monthly basis. After integrating Telegram notification capabilities into their app, the company's overall activity level jumped 34% and revenue increased by 18%. In the finance industry, after obtaining feedback from customers after making a loan decision, AI discovered that customers had commented about not being told the reason that their application was denied. Therefore, when the company began providing customers with detailed explanations as to why they were denied, customer satisfaction (NPS) improved from -40 to -10, and the number of customers returning with a repeat application increased by 28%.

Examples from ASCN.AI Usage: In October of 2023, ASCN.AI was monitoring the crypto market during the flash crash. ASCN.AI was able to identify the price anomalies within hours of occurrence. Traders received notification of these anomalies via Telegram. In less than two hours of receiving this notification, ASCN.AI enabled traders to complete transactions worth over $1.2 million and realize profits as large as 34%. In another example, ASCN.AI detected market manipulation prior to the drop in value of a token belonging to Falcon Finance—two days prior to the drop. Traders who received the notification were able to exit this trade at nearly zero loss.

Automate Processes Across Business Sizes:

Small Business Examples: Google Forms and Google Sheets are free, the use of basic Apps Script from Google, and/or AI from either OpenAI, Claude, etc. ($20 - $50/month). The first 1,000 messages utilizing WhatsApp's Business API are free. Total estimated costs for automation using these technologies would likely range from $30 to $80 per month—compared to costs in excess of $2,000 per month for manual analytics.

Enterprise Examples: Google Workspace Enterprise, Custom AI Models, and/or Corporate Tiers of AI Technology can be used by Large Businesses with integration to a CRM (Customer Relationship Management) and/or a CDP (Customer Data Platform). Total Automation Cost based on processing of up to 200,000 customer reviews: $500 - $2,000. Automated solutions are still much less expensive than employing an entire analytics team to produce analysis on the organization's data.

Technical Setup and Step-By-Step Instructions

Utilizing Forms to Collect Feedback.

Successful feedback should start by creating a form. You can use a straightforward form with 3 fields for the majority of situations i.e. the rating, comment and contact fields.

1. Log in to your Google account and navigate to Google Forms (https://forms.google.com).
2. Create a new form.
3. Assign the new form a title such as "Rate Our Service".
4. Create a Linear Scale question ranging from 1 (not satisfied) to 10 (completely satisfied) and ensure that it is marked as required.
5. Include an optional text field for the comments.
6. Include an email or phone number field that validates the input.
7. Establish conditional transitions based on the selected ratings.
8. Change the appearance of the form to include your own brand colours and logo.
9. Connect your new form to a Google Sheet such that responses entered in that form populate the Google Sheet in real-time.
10. After connecting your Google Sheet to your form, you will receive a URL link to share with your customers.

Frequently Asked Questions (FAQ)

How do you secure data during the automation process? Secure access to your Google Sheets by limiting the number of individuals who have access to edit (aka those with edit permission). Use Multi-Factor Authentication (MFA) to increase security while accessing the data. API keys and tokens should be stored in secure vaults (using the "Secrets" feature in Apps Script). When transferring data, always use HTTPS protocol. If you process sensitive medical or personal data, consider AI services that have obtained GDPR and HIPAA qualified status.

Can this Automation process be used across other business sectors? Yes, the basic theme of the process "Form → AI → Spreadsheet → Messengers" can be adapted for other business sectors, all that will change is the questions you ask, the priority of the questions being asked and the type of integrations you have. For example, in E-commerce you might focus on delivery; in SaaS, UI/UX; in Fintech, processing speed and Crypto, data accuracy and Exchange Integration.

What is a solution to Typical Issues that arise?

• When Rate Limit is Exceeded: Apply a 1-second delay between API requests or upgrade to an enterprise tier.
• When Your Trigger does not Fire: Ensure that your trigger settings and limits have been set correctly. Review the trigger and optimize your code.
• When the WhatsApp API does not send: Ensure that the templates you are using are approved and your phone number is verified in Meta.
• When AI Misinterprets Sentiment: Refine your prompts, include sample input, perform fine-tuning.
• When Google Sheets is Slow: Reduce old data, Optimize formulas, and remove unused calculations.

In Conclusion – How Automation can aid in Improving Customer Support.

To summarise, Feedback Automation not only improves response times but also shifts to a proactive approach. When using a machine, there are no days off - thus capturing subtleties/nuances that may not be noticed by an individual. While Automation does not replace the need for human contact, it allows your team to concentrate on their actual responsibilities rather than repetitive tasks. It is always advisable to begin with the smallest of changes: create an initial form for one main Milestone/Area, connect a simple AI to it, and implement a Messaging Platform that facilitates feedback. Within a week, your customers will notice a significant reduction in the time it takes for your organization to respond, they will have improved satisfaction and view that feedback will allow you to obtain insights themselves. Subsequently, you can expand the scope of your Automation by adding additional feeds (data sources) to your Automations. Finalize your AI Enhancement level and incorporate your CRM into Automation over a 3-month period, your automation will become an invaluable member of your team.