
A new startup building a new product is like an untrained sniper trying to hit a bull's eye from a mile away.  You are likely to miss the target.  Luckily, you have the superpower to change the trajectory of a bullet mid-flight.

That superpower is called a customer feedback loop.

## What is a customer feedback loop?

A customer feedback loop is a process that involves collecting customer feedback, analyzing it and then implementing changes based on the collected feedback.

## Why should you use a customer feedback loop?

Feedback helps you build a relevant product, enhance customer relationships and make informed decisions.  The loop lets you stay on top of what your customers think and how they react to your product.

## When should you start a feedback loop?

As early as possible, usually at the proof-of-concept stage and throughout product development.  If you are developing a minimum viable product version without a feedback loop, you are increasing the risk that the product will not meet the customer's need.

## How often should you get customer feedback?

With the new product, you should use a short feedback loop measured in weeks, allowing you to focus on relevant features and pivot quickly.  As the product matures, you can increase the length of the loop.

## How can you get customer feedback?

* Send customer surveys about product features, usability, customer service, and overall satisfaction.
* Monitor online reviews and ratings.
* Track conversations and interactions in public forums and compare those with your competitors.
* Conduct customer focus groups where you can directly talk to customers to get detailed feedback on specific features or areas of improvement.
* Display a customer feedback form or a chat on your website or as part of your product to make it easy for customers to provide feedback.

Prioritize customer feedback loops for any new product or feature you develop, thus curving your product bullets towards your targets.



How to Curve Product Bullets with Customer Feedback Loops


Trust will make or break your startup.

Trust is what closes the deals and makes customers buy your products. Trust is why people return to the same stores or buy the brand. People trust the brand based on their previous experiences with that brand's products and services.

Your startup level of trust with your customers will determine your ability to acquire, retain, and monetize them.

As the saying goes, you cannot manage what you cannot measure. So how do you measure customer trust?

Customer trust has six measurable components: satisfaction, sentiment, loyalty, churn, virality, and value.

First of all, your startup should focus on customer satisfaction. Customer satisfaction influences customer sentiment, loyalty, churn, and virality. You can measure customer satisfaction via surveys and feedback, product and customer service ratings, and response times.

Your customer sentiment is your brand reputation and influences your customer virality. The sentiment is cumulative of your reviews, social media comments, and google searches. You can measure your customer sentiment via social media channels and how others speak about your company. You can also calculate the share of voice to determine what percentage of people are talking about your brand as opposed to your competitors.

Customers you gain trust with are much more likely to engage with your company and repurchase your products.  You can calculate the repeat purchase rate to measure customer loyalty or how many customers repeatedly buy your products and services. 
 
Also, keep track of your customer churn. Your customers churn because they lose trust in you. A startup could cause churn by promising too much and not delivering, thus causing a mismatch between customer expectations and reality.

Measure customer virality using Net Promoter Score (NPS), and determine how quickly your business grows via word of mouth. NPS is a simple tool that asks customers to evaluate how likely they are to recommend your product or service to others. Viral cycle time, or how long a user invites another potential user onto your product or service, is one of the best indicators of your growth. A high NPS will decrease your viral cycle time.

Customer loyalty, churn, and virality will influence your startup's profitability. One of the most critical measures of a successful startup is the Lifetime Value of Customer (LTV). Use LTV to identify high-value customers and market segments. If you acquire and keep customers cheaply, you will be in a great position to grow your business.

A startup customer journey starts with customer satisfaction and ends with monetization. Use customer satisfaction, sentiment, loyalty, churn, virality, and value as your compass.


How to Measure Customer Trust from Satisfaction to Value


We hear the terms POC (Proof of Concept) and MVP (Minimum Viable Product) used interchangeably daily. When a startup seeks to build a product and engages a third party, this ambiguity could misalign expectations between a startup and a service provider. So what is the difference between POC and MVP, and when would you build each?

Startups use POC to test the feasibility of their idea with a few potential customers when the startup is unsure if their idea works. POC is not built for scale but only for those few initial users. You would not use POC to discover PMF (product market fit) since PMF is a degree to which a product satisfies customers in numbers.  

POC usually takes days to weeks to build and is an excellent option for bootstrapped startups to validate a business idea as efficiently as possible. POC is more of an exercise than a product and does not scale beyond a few users. As a bootstrapped startup, building an extensive product before getting feedback from the first few users means spending too much time and money on something that might have little to no use.

POC also does not have to be automated or be software at all. For example, if you are building the next Amazon, your first goal should be to move a few packages from a seller to a buyer. You would use people to package, deliver, and send notifications without always being concerned with automation. If you managed to satisfy those first users of the system and demonstrated your idea's differentiators, you might have completed a POC and are ready to build an MVP.

To measure the success of POC and validate your idea, you would get feedback from those users participating in the exercise through surveys and interviews.

Once you have already validated your idea and proven it works, congrats! You might be ready to build your MVP. Now, you are ready to automate and scale your core processes and functionality. The goal for MVP is to discover Product Market Fit and be able to pivot, test, and iterate with your product.  

MVP usually takes several months to build and is used by bootstrapped or funded startups to find PMF. Since finding PMF involves validating with customers in numbers, your MVP should be a quality product that scales to support the increasing number of simultaneous users.

Unfortunately, many MVP implementations sacrifice product quality. That's the wrong approach. The critical letter in MVP is V, which stands for viable. MVP saves effort vs mature products by providing a minimum number of polished features and should be built to the standards of commercial products. The subpar-quality product would block user growth and prevent PMF discovery.

To measure MVP's success, you want to go beyond gathering direct feedback from users via surveys and interviews. You want to analyze customer reviews and track user analytics, such as viral cycle time or the amount of time it takes a user to invite another user to your product or platform.

|| POC | MVP |
|-| --- | --- |
| Why build it? | To validate your idea | To find product market fit |
| What is it? | An exercise focused on testing if your idea can be turned to reality | A usable quality product with a minimal set of features |
| Build to scale with users? | No | Yes |
| How long to build? | Usually Days or Weeks | Usually Several Months |
| When to build? | Bootstrapped startups with zero to few users | Bootstrapped or funded startups with few to many users or customers |
| How to measure success? | Surveys, direct user feedback, interviews | Surveys, direct user feedback, interviews, customer reviews, viral cycle time, user analytics |


POC vs MVP


### Amelia’s Journey

Hi, I am Amelia, and I heard about this fantastic web platform that your startup has developed. I open my Macbook and type in your web address. Your website tells me it can’t be reached and gives some unintelligible error about the DNS probe. I try again and get the same message. I am going to tell my friends that your platform does not work.

Your tech-wiz friend later tells you that he made a small error in the NS configuration. It doesn’t mean much to you, and they say it’s no big deal, and they fixed it, so it’s ok.

Or, something popped up on your newsfeed about the latest DNS flood attack bringing down DNS Resolvers in the western part of the US, but boring, and who cares anyways.

### Sean’s Journey

Hi, I am Sean! I am about to go check out your website. I alt-tab from Call of Duty and open your site in another tab because I don’t want to close live streams I watch on another monitor. Great logo and graphics! I scroll down, but it is choppy. I click a button, and nothing happens. I click 1000 more times to be sure. This site sucks! Alt-tab back to the game.

Your web dev said something last week about optimizing your uncompressed images and videos due to their memory usage and blocking javascript. But hey, that’s not important since the website already looks good, and you need to ship a feature that your competitor released last week.

### Giorgia’s Journey

Ciao, I am Giorgia from Genoa, Italy! I heard about your excellent website. I type in your address in my browser. I see something loading… still loading… still loading… Ciao!

Your dev mentioned something about using content delivery networks more effectively and deploying to another region to improve latency. She said it was going to take two weeks. That’s two weeks too long, and the site loads fast on your laptop, so nope.

### Sanjeev's Journey

Hi, I am Sanjeev from Mumbai, and we need that enterprise tooling from your website for $999. We fill out your order form with our address and excitedly click submit. Your site surprises us with 400 error. We need that tool now, so we are buying it from your competitor.

Last week, your devs rolled out the backend address validation library since you got a lot of order spam. They tested it in US and Canada, and it works great! Particularly proud of the regex that passes on only Zip and Postal codes and fails on everything else.

### Karen's Journey

Hi, I am Karen from Indianapolis! Your product is gold! I tab tab tab through your order form and click submit. Nothing happens. I click again. Nothing happens. I look up, and one of the fields is red. I guess I somehow skipped it. Oh well, I'll go back up and fill it out. But wait, my moms calling me, so I will do it later.

Two hours later: what was the name of your site again?

Your old dev left, and you asked your new dev to quickly add a product field that should only take five minutes. He did it quickly, in five minutes and deployed the new site. What's a tab order anyways?

### Bob's Journey

Hi, I am Bob from Seattle. I am on the go and order a bunch of your products on my cell. I put in my credit card information. Darn it, that card is expired! Ok, pulling out the new one…. What happened to my order???!

Last week you overheard your devs arguing about ephemeral vs persistent state and potential data loss. So you asked them to keep the argument ephemeral and go back to coding. So they did.

### Marina's Journey

Hi, I am Marina from Kyiv! I am in the coffee shop, and I found an open WiFi! I go to your website, enter my username and password, and log in. I order a bunch of stuff and put in my credit card information. I love your site!

Next week: My credit card is blocked. :( Who ordered all those purchases for $10,000?

Your operations folk claim your site is secure and show you an encrypted end-to-end ordering process. They uploaded the certificate to your server. The site works so well that users can access it via HTTPS and HTTP.

### Cheng's Journey

I am Cheng from Vancouver! I am using your service to stream how-to videos from the library PC. The library is closing, so I've got to run.

Two months later: I didn't watch any of those videos. Why am I getting those usage upcharges?

Your engineers quickly implemented token-based authentication. The design's key features are the storage of tokens in persistent storage without an expiration date.

### Black Hat's Journey

Hi, my name is unimportant, and neither is my location. I go to your website and signup. Your site returns ?customer_id=5000. Then, I go to ?customer_id = 4999, 4998, 4997… My pet bot got the emails and addresses of all your customers.

Your junior dev shipped authentication. Only authentication.

### Annette’s Journey

Salut, I am Annette from Lyons! I integrated my client with your API, and the client is fetching critical data for my application. Now, I get calls from customers saying my app does not work. The API calls take over 20 seconds or time out. I am losing customers!

Your devs quickly wrote queries for your single-instance SQL database. The queries work, even without indices, JOINs, or LIMITs.

### Mike’s Journey

Hey, I am Mike from Toronto! My app now uses your API. What a terrific product!

A month later: My app is broken. What’s wrong with your API?

Your dev released all kinds of goodies and features. All are deployed to the one API version to rule them all. Who needs contracts, anyways?

### Emma’s Journey

I am Emma from Philly! I integrated my open-source app with your API. My engineer created an API integrator module with production API credentials and pushed the module and credentials to my public repo.

A week later: You are charging me $20,000 for usage?!!

Your API request logs have all kinds of funny IP addresses. Emma from Philly issued millions of requests from all sorts of locations. However, your devs are working on features, and your single operations guy is burning the midnight oil, just keeping the lights on.



A Startup Customer (Mis)Adventures Through The Web 


### Validating Market Idea with Serverless
To validate your market idea, you want to develop and deploy your prototype as soon as possible and iterate on it. When building a proof of concept or minimally viable version of your product, the last thing you want is to be slowed down by infrastructure and operations. Worry about that after you validate your idea and have paying customers.

The serverless model has transformed software development by eliminating the need to manage your application infrastructure. Serverless leverages your cloud provider data center capabilities to manage your infrastructure. Adopting a serverless model means faster application development and deployment while leaving scaling and securing your infrastructure to your cloud partner.

According to the State of Serverless studies done by Datadog in 2021 and 2022, over 70 percent of organizations running in AWS cloud adopted serverless functionality.

The two frameworks that help the development of serverless infrastructure are Serverless by Serverless Inc. and Serverless Application Model (SAM) by Amazon Web Services (AWS). These frameworks enable you to deploy and get started with your application in minutes.

Both Serverless and SAM act as shorthand for complex declarative infrastructure templates. In addition, they give you CLI tools that allow you to deploy and manage your application from the command line.

Serverless supports different cloud platforms, such as Google Cloud and Microsoft Azure Cloud, in addition to Amazon Web Services (AWS), and gives a web-accessible UI and dashboards.

AWS SAM provides a local Lambda-like execution environment and integrates with AWS SAR (Serverless Application Repository) for publishing serverless applications in AWS.

### AWS Lambda Under the Hood
AWS Lambda is Amazon's implementation of serverless computing as an alternative to a container or virtual machine infrastructure.

*What can AWS Lambda do?*

We ran AWS Lambda functions for a wide variety of use cases to do the following:

* Backends for web and mobile applications
* REST and GraphQL APIs
* Queue producers and consumers
* Data transformation and processing
* Authentication and authorization
* Near real-time analytics
* Scheduled cron jobs
* CI/CD pipeline monitoring

*How does AWS Lambda work?*

AWS Lambda runs on Firecracker, a virtual machine monitor (VMM) that uses the Linux Kernel-based Virtual Machine (KVM) to create and manage microVMs.

KVM controls the host CPU to perform hardware virtualization, memory management, and hardware abstraction.

Firecracker sits in the host OS on the IO path, configures KVM, and emulates hardware. Firecracker is optimized explicitly for serverless with 5-10MB overhead, can boot Linux in 125ms, and supports the creation of up to 150 microVMs per second per host.

Thus, Firecracker enables running thousands of Lambda functions on the same hardware without having thousands of network cards or SSD drives.

Here's what the AWS Lambda stack looks like:

1. Code
2. AWS Lambda
3. AWS Lambda Runtime
4. Guest Kernel (Minified Linux)
5. Firecracker / KVM
6. Host Kernel
7. EC2 Instance

### AWS Lambda Deployment
*Do you deploy bugs to production and take too much time to recover?*

You can solve this problem using serverless AWS Lambda by leveraging the functions' built-in versioning and aliasing. A Lambda alias points to a function version, and you can update the alias to point to a new version of the function after deployment. Thus, AWS Lambda allows you to create and manage multiple versions of your function code.

Each AWS Lambda alias has a unique ARN or Amazon Resource Name. Other system components in your cloud system architecture reference this alias ARN. Thus, if you configure a production alias and change the underlying version of Lambda, you don't need to update the other system components, saving you time and effort.

Thus, you can leverage Lambda versions and aliases to deploy safely and recover fast from production defects.

### Running Data Intensive Workloads on AWS Lambda
*You can start running your machine learning or data-intensive workloads in the AWS cloud in no time on serverless infrastructure.*

The standard way to upload a code into AWS Serverless Lambda Function is to create a .zip file. Unfortunately, Lambda has a frustrating 50 MB .zip file limit that can prevent data scientists and engineers from using it to deploy their workloads, which usually reference large libraries and can grow to gigabytes in size.

Luckily, you can package and deploy your containers on AWS Lambda. Using a container image with Lambda can be as simple as fetching an AWS base image and creating a Dockerfile with your dependencies installed on it. This approach enables you to deploy applications up to 10 GB in size and supports most machine-learning and data-intensive applications.



Our thoughts on technology, leadership, cloud, software, and more

How to Curve Product Bullets with Customer Feedback Loops

How to Measure Customer Trust from Satisfaction to Value

POC vs MVP

A Startup Customer (Mis)Adventures Through The Web

Amelia’s Journey

Going Serverless with AWS Lambda

Validating Market Idea with Serverless