GROWTH: SIGNUP & ONBOARDING
Arc.dev — Optimize Sign Up Flow & Onboarding (Dec, 2020)
Proposed Experiments to Improve Onboarding & Activation Metrics
About Arc.dev
A hiring platform, powered by Codementor — a mentorship platform to get help from vetted software developers — helping leading tech companies and startups that aim to efficiently grow remote teams, recruiting quality vetted developers from all around the world.
In other words, it is a strategic product/service to monetize vetted developers acquired from their mentorship platform, who are open to work remotely, by placing them as an temporary or permanent expansion, or to fulfill short-term projects for much smaller scaled companies.
Outline
- Business Model
- Value Propositions
- Problem Statement
- Process
1. Destruct Sign-Up Flow
2. Identify Frictions
3. Define Activation Funnel & Primary Moments
4. Propose Experiments:
#1 Create More CTAs That Triggers Sign Up Flow
#2 Guide Users Enter Vetting Process ASAP
#3 Validate Email in Real-Time (+ Email Verification)
#4 Specify Unfinished Steps for Sign Up Flow Process
#5 Emphasize Values-Added Throughout Sign Up Flow
#6 Reduce Steps & Delay Custom Inputs Later Down In The Process
#7 Display Vetted Badges on the User Profiles
#8 Provide An Integrated Dashboard
#9 Improve Activation by Supporting Multi-Tag Based Searching
#10 Improve Activation by Generating More Personalized / Custom-Filtered Job & Project Recommendations
Business Model
This business, recruiting, currently runs on a transactional model, which means they earn only from successful placements, either a fixed fee on permanent full time hires, or a % of all logged hours on contract hires — a classic two-sided market: employers (B, D) and candidates (A, C).
Another way to increase Revenue is by increasing Fixed Fee or Commission %, both of which will depend on bargaining leverage of the Business Development Team with their clients (employers).
The more quickly Arc.dev can fulfill successful placements for the clients, the more drastically their revenue can grow.
Value Propositions
To incentivize both sides to adapt their services, as subtly addressed on their website, which can also be found from a Competitor Analysis + going through their Client Testimonials, the value propositions that Arc.dev offers are:
For Employers — Demand (B, D)
- Save time and effort screening for qualified candidates.
- Find pre-vetted developers ready to start immediately.
*Full-time ≻ Contract
For Developers — Supply (A, C)
- Search all remote developer jobs in one place with custom filters.
- Receive priority matches by email shortly when remote companies post a new position.
- Get pre-qualified and fast-track to final interviews.
*Applicable for verified jobs
Problem Statement
In this case, finding great developers appears to be harder than finding people who want to hire them, and I assume Arc.dev has validated buy-in from the “hard” side by stating buy-in from the “less hard” side.
Hence, I’m going to tap into addressing the JTBD (Jobs To Be Done) that can help growing the supply of developers based on their value propositions, preferably the full-time for a stronger demand and higher weighted value to the revenue.
Business Objective: Grow The Supply of Developers
Assuming the Product Marketing Team succeeds in effectively directing high-quality leads across all channels to Remote Job Site, it’d be likely to help realize the goal from the Product side by:
Primary Goal — Optimize Onboarding Efficiency
Then, ideally feeding insights learned from experiments back to the Marketing who can tailor Acquisition Strategies to:
Secondary Goal — Increase Amount of Quality Leads to Top of the Funnel
Process
It’s a complicated product considering its business processes, so I will focus solely on improving “Signup” & “Activation” for developers on Remote Jobs Site by Arc.dev in this case study — one of the primary metrics in the company.
I first start from analyzing their Sign Up Flow and Activation Funnel to understand the efficiency of onboarding developers,
, and then create experiments the might help improving the metrics.
Metrics
- # of Sign-Ups ≂ # of Visitors Who Submit Emails
- % of Sign Up Conversion ⥮ % of Sing Up Drop-Off
- Click-Through % & Drop-Off % at Each Step of the Sign Up Flow
Destruct Sign-Up Flow
How Many Unique CTAs (Call-To-Action) Are There on Arc.dev Remote Jobs Site?
There are totally 4 CTAs throughout the Remote Jobs Site, 1 on the header (left), 2 below each position listing and Company Brief (upper right), and 1 just above the footer (lower right).
# of CTAs: 4
# of Unique CTAs: 2
2 Unique CTAs: Sign Up (Now), Begin Smart Search
What Triggers Sign-Up Flow?
All CTAs on the Remote Job Page do. However, there are two different paths that would eventually lead to trigger the same Sign Up Flow:
1. “Sign Up (Now)” ➝ Sign Up Flow
2. “Begin Smart Search” ➝ Set Custom Filters Flow ⇢ Sign Up Flow
Type of Sign-Up Flow
Friction Based Service Signup Flow
Arc.dev ask for Email, Full name and Password here, and there’s going to be a multi-step form (as shown in Arc.dev Remote Jobs Sign-Up Flow). This would be a Friction Based Service Signup Flow, because they are collecting additional information to either personalize or enhance the user experience.
Furthermore, telling from what Arc.dev advertise themselves on their Remote Job Site, they create a personalized list of Software Developer job vacancies that match users’ expertises, technologies, goals, convenient timezones, and countries (custom inputs), to save the hassles from users of spending time scanning for suitable positions.
What’s Friction Based Sign-up Flow?
A users flow that makes it more challenging to complete the sign up. By doing so, the users are getting additional value in the future.
Identify Core & Custom Inputs
First, they ask for simply their core fields. Once a user signs up, it comes to the second part — custom fields.
# of Core Fields: 3
# of Custom Fields: 12
Think: What custom fields does Arc.dev business really need?
Order of Operation
Arc.dev ask for core inputs first and then custom inputs. What further helps conversions is that they segment the core and custom inputs into a 8-step multi-step form. The order of operation does flow correctly.
Calculate The Number of Fields & Clicks
Be aware of every click and every field as each step counts. Even if just a click in the box field to start typing, that is a click itself.
# of Clicks to complete: min. 23~40+
# of Fields to address: 15
Identify Frictions
Since I don’t have access to the numbers, like Conversion % and Drop-Off %, to understand how efficient the current sign-up flow is,
Friction: The psychological or digital resistance that visitors experience when trying to complete an action, where users might have to pause, think more, or slow down their flow to get to their goals during the process.
I then start from inspecting for frictions, as a user that first come to Remote Jobs Site and enter the Sign Up Flow:
Good Friction
Pre-populated Time Zone field by using reverse IP
The user will have a better experience as it’ll show that they understand who the user is and where they come from. It might also help build trust that the company has worked with others from their countries.
About accuracy? The time zone might not always be correct. Let’s assume that 90% of the time if will work. For the other 10%, they also allow users to change the time zone.
Think: What happens with trust when they auto-select the wrong time zone?
Bad Frictions
Bad friction results in higher drop-off % and lower conversion %. To name a couple examples…
Overwhelming Steps & Fields to Complete Sign-Up Flow
It takes at least 23 clicks (up to 40+ clicks if completing all custom fields) to go through 15 Core & Custom fields in a 8-step multi-step form to finish the Sign-Up Flow.
To users, the #1 expectation is to test the product or service, not to fill out a bunch of information. If the intention is to ask users to complete their User Profile, maybe making it separate from the Sign Up Flow and collect those custom fields later down in the process.
Ask: Is it really necessary to ask users for every field during the sign-up?
Hard to Find Positions That Suit Users’ Interests The Best
Say, if I were a Senior Big Data Engineer with 5 years of experience, interested in exploring remote job opportunities, then I’d be a visitor with one specific intent that I’m looking for remote positions that require my tech stacks or relevant projects that I’ve once participated.
However, the search bar allows only one keyword along with “job level (seniority)” and “job type (perm/contract)” filters to custom search results, which gives still a relatively wide range of outcomes. If I am not able to quickly find something highly related to what I’m looking for, I might just bounce to other sites that supports more powerful searching.
Scary Fact: If a user bounces…he (she) might never come back.
Define Activation Funnel — Primary Moments
We can improve Activation by…
- Improving % of Conversions through each moment in the funnel
- Bringing forward the Aha Moments, which should echo back to their Value Propositions:
So that users can experience the values before the Setup Moments (reducing the time-to-value).
Sub-Metrics
How long does it need to take users to complete each primary moment? (assuming the ideal case is done all within 14 days after sign-up)
3 steps to Setup; +4, Aha
Segment Analysis
Another way to improve activation is to make improvements for different segments of users — demographic, firmographics, or product usage.
Segments include: Time zone, tech stack, seniority, age, what they’ve done in the products, where they come from, where the lead came from, etc.
Advanced example (Cohorts): Users who complete vetting tasks within 1 week — less time than average.
Moreover, by leveraging the understanding for different User Segments to send the right emails at certain moments during onboarding, it will also benefit increasing conversions at each stage and overall product engagement.
Sub-Metric
- The delta in between each moment → The # of users that have reached Setup moment are actually reaching the Aha moment; Aha, Habit
- The # of users activated (going through the entire funnel)
- The % of users activated
- Which segments are performing the best and the worst
- Average of all attributes within each segment perform with each one of these metrics
Propose Experiments
First, I…
❒ Experiment Framework
- Define the goals — aligning initiatives with business rules
- Ask narrowed drill-down questions to understand the “why”
- Hypothesize what will happen — impacts on the metrics of other stockholders/teams
- Prioritize for low efforts and high impacts: ICE Framework
- (won’t be able to for now, but should) Test & iterate
Then create Hypothesis-Based Experiments, using:
❒ Experiment Brief
[Action], so that [Outcome], and because [Theory]
[Action] = what I want to test
[Outcome] = what I expect will happen (impacts on the metrics)
[Theory] = why I think the action will lead to the outcome (assumptions, resources, references from previous experiments)
The business objective for Arc.dev is to grow the supply of developers.
Primary Goal: Optimize Onboarding Efficiency
Secondary Goal: Increase the Amount of Leads to Top of the Funnel
Hence, experiments include:
#1 Create More CTAs That Triggers Sign Up Flow
[Action] In addition to the current four “Sign Up (Now)” & “Begin Smart Search” CTAs, create other ideal signup paths with forms that help an unknown user convert to a known user, including:
- Email Signup Form — for Newsletter & Blog about Remote Jobs and user-generated content about recent technologies, etc.
- Purchase Signup Form — when buying access to Live 1:1 Mentorship Session for code review or interview preparation (engineers) / vetting service from experienced mentors (companies)
- Subscription Signup Form — when creating an account to join Arc.dev Developer Community for Q&A -OR- to receive weekly/daily email of new “custom filtered” remote jobs, ex: “Product Manager”, “Design”
Moreover, place these additional CTAs as primary to get visitors on the website to go through work flow and then asking them to sign up.
At last, measure the conversions through these flows, by calculating the click-through rate and the drop-off % at each point or stop, and later compare to find out what is the most efficient way to generate leads.
[Outcome] We can increase Sign Up Conversion % by increasing the amount of leads to the top of the funnel
[Theory] By casting a wide net for all the different types of visitors Arc.dev might receive on the website, it’s likely to collect a longer name list with contact info (email).
#2 Guide Users Enter Vetting Process ASAP
[Action] Create “Take Quick Quiz” CTA on the homepage as a shortcut for users to kick off Vetting Process before the Setup Moments. Moreover:
- “Get Vetted”or CTA right in a full-page popup window after users click “Apply” on a Job Page
- “Get Vetted for Free” to Fast Track to Interview CTA as a clickable link on every job listing
[Outcome] We can increase Sign Up Conversion % by increasing sign-ups
[Theory] In order to fulfill the value propositions that Arc.dev aims to attract clients to adopt their services, the key is to source as many vetted developers as possible, ASAP, so that Arc.dev will have a growing pool of qualified candidates ready to place whenever there’s a demand from companies.
#3 Validate Email in Real-Time (+ Email Verification)
[Action] Immediately verify whether users input a valid email, moreover, prevent them from advancing to the next steps if the validation is still in process.
[Outcome] We can expect increasing Sign Up Drop-Off % by generating more valid sign-ups / increasing accuracy of user profile data, whereas there is also a possibility that users might choose to quit the Sign Up Flow if we don’t allow users skip email validation.
[Theory] A done analysis shows that leads with accurate emails convert more, and the Arc Marketing Teams would be able to re-target them later even if the user doesn’t end up registering with Arc.dev.
#4 Specify Unfinished Steps for Sign Up Flow Process
[Action] Show users how many incomplete step they’re going to address before reaching the end, for example: 3/8 (as done in the Set Custom Filter Flow), to indicate that there are 5 more steps before they finish this multi-step form.
[Outcome] We can decrease Sign Up Drop-Off %
[Theory] Not only will the users feel they’ve already made progress, they also know where they are in the process and how much work is left to do. It also utilizes the endowed progress effect.
#5 Emphasize Values-Added Throughout Sign Up Flow
[Action] When collecting the info from these use cases, mention it is meant to personalize the experience for user based on users’ feedback — this is the value to the users.
[Outcome] We can decrease Sign Up Drop-Off %
[Theory] So that users know the value they will receive, instead of perceiving they are only providing what the company needs.
#6 Reduce Steps & Delay Custom Inputs Later Down In The Process
[Action] Remove unnecessary steps & fields, for example “Upload Profile Picture”, from the Sign Up Flow and delay this input until when users finish vetting and their profiles are close to 100% comprehensiveness to apply for verified jobs. Moreover, separate Configure User Profile process completely from the Sign Up Flow.
[Outcome] We can decrease Sign Up Drop-Off %
[Theory] Forcing a user to choose an avatar or upload a headshot photo before getting into the product creates friction. Also, for a complicated product as Arv.dev, where users need to go through many different processes to be qualified for fast tracking to the final interviews, trying to onboard users in a single sitting can be overwhelming and actually reduce motivation.
#7 Display Vetted Badges on the User Profiles
[Action] Create badges that can indicate users’ achievements on different stages of the Vetting Process, that will display automatically on the User Profile as soon as a user finish a requirement.
[Theory] The more engineers use Codementor to up-skill themselves, the more badges they get, and the better their profile looks attractive to potential employers. Everyone likes the idea of leveling up, and this is a great way of tying personal development into prolonged use of the product.
#8 Provide An Integrated Dashboard
[Action] Set up milestones and offer incentives like “missions / tasks” to make the vetting process more interactive along the way.
[Theory] This encourages users to take the next step and continue their contract to finish the comprehensive Onboarding flows.
(To be continued)
Moreover, one way to decrease churn is to establish perceived value propositions of Arc.dev to current user at an early stage,
Hence:
#9 Improve Activation by Supporting Multi-Tag Based Searching
[Outcome] We can increase leads generated on Remote Jobs Site itself by removing the fiction while users exploring desired positions + by gathering clues to create new touch points to reach the users. By doing so, we can improve conversions down the funnel by guiding prospects to the Aha moment before signing up.
[Theory] By using a tag based search bar that allows multiple inputs, users will be able to share their intent of why they come to the website and the platform would be able to collect the intent of why users are visiting. Even if the user doesn’t end up registering with Arc.dev, the team would be able to re-target them later. A win-win situation when removing the friction, and to get users to experience one of the Arc.dev’s core value as fast as possible.
Tag based search: The ability to select tags and perform a search could simply be called. This is a form of faceted search.
#10 Improve Activation by Generating More Personalized / Custom-Filtered Job & Project Recommendations
[Outcome] By doing so, we can improve conversions down the funnel by decreasing the time-to-value to guide prospects to the Aha moment.
[Action] Based on Product Usage collected on Codementor, send emails not only to Arc.dev users but also Codementor users that we know enough to personalize a job list based on their tech stacks, preferences, etc. Temp with a pop-up dialog while in action or with an email to ask for more information about them if not enough inputs.
Further Action
- Get buy-ins from stakeholders to execute experiments
Some Extra Ideas…
Inbound Supply Growth Engine
(To be continued)
