I am officially a Survey Gizmo believer. After struggling with a dozen survey-builders, I finally found my solution in Survey Gizmo ($50/mo).
The Sprint:
I started with a spreadsheet with data on what each client liked + their budget (i.e. Abstract art, $500 & below).
Survey Gizmo pulls from my spreadsheet and shows/hides artwork based on what the user’s preferences. We can then send each user a custom link with their preferences pre-populated.
It looks like this:
2) ‘Liked because’ feature
Context:
Lately, my friends have been talking about a dating app, called Coffee meets Bagel.
Coffee meet Bagel show users a daily match recommendation. Users rate the match.
Then, the app is smart enough to ask WHY the user liked or didn’t like the person. Which is super smart, because most other apps just infer why you liked someone.
The Sprint:
Built a similar “Liked because” feature to gather better data about what people like so that our curators can send better recommendations.
Learnings:
I’ll come back to this post & add my learnings since this just launched today!
For my last sprint, everyone saw art in all categories (regardless of what they subscribed to). The mission of this sprint was to start serving different combinations of art recommendations to different people…without coding.
The Sprint:
Campaign Monitor ($29/mo) has a really awesome feature, called “Dynamic Content”. It allowed me to only email people with art in the categories they opted into.
Here’s a demo of how it works:
2) More info option:
Context:
My curators spend a lot of time finding art for clients.
The most time consuming part is writing paragraphs on each piece of artwork… that most of our clients never read.
So we decluttered the art recommendation pages & started linking to a separate page where clients can get more info on pieces- only if they want to!
Learning:
I didn’t really have my sh*t together for this sprint.
I was able to show each user a different combination of artworks in their email. But when users clicked out to the Google Survey, they still saw every piece of art that was curated for the week.
Google surveys doesn’t have a code-free show/hide feature. So, again, I was worried about this.
But launching anyway, confirmed my learning from the first product sprint.
A bunch of people still found art they liked (rated 4 or 5) in categories they didn’t sign up for.
Note to self: Don’t spend a ton of money down the road on an advanced algorithm to curate art. This targeted service doesn’t need to be perfectly targeted
I made a big change to Kollecto to make it scalable for the hundreds of active users we now have.
In a lot of ways it felt like I was starting everything from scratch- moving from a 1:1 advisory model & to a personalized, but scalable (1: many) model.
Last month, I pushed myself to do regular ‘codeless product sprints’.
I’ve done three major ones so far and am now I’m committing to doing small ‘codeless product sprint’ each week.
I’m using this blog as a way to hold myself accountable! 🙂 So I’ll be writing weekly about new features I’m adding to Kollecto. Plus I’ll be documenting the lessons learned from getting hundreds or users to engage with the features.
New features:
Art categories
User Ratings (for each artwork)
1) Art Categories:
Context:
For the past 6 months, we’ve been asking open-endedquestions about what kind of art people like/ want.
There were problems with this:
A) most people didn’t have the ‘art vocabulary’ to tell us what they like.
B) The process didn’tallow us to easily group people with similar taste.
The sprint:
Introduced art categories that people could opt into [using Typeform ($25/mo)]
2) Ratings:
Context:
This is actually something we used to do.
We stopped allowing clients to rate artwork because it required making a new client presentation for each user!!
But as we intoduced scalable categories, we added the ability to rate each piece again!
The sprint:
Created a Google Form to show art in each category & let clients rate the pieces. Each client got an email like this.
Learnings:
For this sprint, everyone saw art in all categories (regardless of what they subscribed to)
At first I was worried about this (I hadn’t yet figured out how to easily serve different combinations of art recommendations)
But it turned out that lots of people found art they liked in categories they didn’t subscribe to!
Learning of the Sprint= By accident, I learned that this targeted service doesn’t need to be perfectly targeted. Sometimes it’s best to include some suprises in your product.
I’ve honestly lost track of all the ways we’ve iterated on the Kollecto UX.
It’s been an awesome journey that feels a lot like putting together a difficult puzzle.
.
.
Here’s a log of all the UX iterations we’ve done to date…
.
SIGN UP
Showed clients 4 pieces of art gauge their taste(discontinued because the results weren’t very helpful to advisors)
Asked clients to share URL to an example of art they like(discontinued because it didn’t depict clients’ multi-dimensional taste)
Asked open-ended question about what the client’s looking for(continues to be helpful indicator on how to help clients- even if they can’t fully articulate what they want)
Asked whether clients have bought art before(continues to be helpful indicator on how to help clients)
Asked clients what mediums they were interested in (i.e. painting, drawing, etc) (discontinued because most of our clients don’t know what they like yet. Most people selected all mediums!)
Asked clients to pick a number on a scale- indicating whether they’re decorating or looking to build a ‘serious’ collection(continued to be helpful indicator on how to help clients)
Asked clients what size art they wanted (discontinued because almost everyone thought bigger was better)
.
PRODUCTS/ SERVICES
3 options: Student Art Advisor | Emerging Art Advisor | Aspiring Investor Art Advisor (discountinued for more simple operations)
1 option: Personal Art Advisor(discontinued
Email class on affordable art collecting (continued- the class drives sales & converts users)
2 options: Shared Art Advisor | Personal Art Advisor(continued)
Hidden product: Free Personal Art Advisor trail (can only be accessed via class)
.
PRICING
Let customer choose monthly vs. broker fee(discontinued because 75% clients chose broker fee)
Flat Advisory fee only (continued)
Shared Art Advisor service- free to customer vs Personal Art Advisor service- 10% advisory fee (new)
.
CLIENT COMMUNICATION
Video chat first, then email (discontinued, high interest in scheduling video chat, low follow-through)
Let customers choose, video chat first, then email; phone call first, them email, or email only (discontinued because 67% chose email only)
Email only
Option to schedule phone call after 1-2 emails of art had been sent (reintroduced b/c we have high success with clients we talk to. This went well)
Advisors manage their own client communication through a shared inbox (that can be monitored) (discountinued because each art advisor provided different customer service, it required more work than the small commission)
Full-time Account Managers manage client communication; advisors find & submit art for client (continued)
Find art for all clients who sign up(discountinued, created a lot of work for clients who werent fully interested)
Require clients to respond to Art Advisor’s email– then find art only after a response (discountinued- we lost clients who were looking for art, but didnt want to chat)
Find art for all clients who sign up, but require feedback after two emails/ 10 pieces shown (continues)
.
ART PRESENTATION
Show art using typeforms where clients rate each piece (discontinued for more simple operations)
Show art within email- with links to request more info, links to give feedback (continued for personal art advisor program- open ended questiond actually give us more feedback than ratings)
Show art within email- with rating (continues for shared art advisor program)
Instead, we’ve builtour MVP by stringing together off-the-shelf technology.
As a non-technical founder, building without code has given me full flexibility to iterate on the Kollecto UX without spending pre-mature time or money on tech. It allows us to spend under $500/mo. on all our product, marketing, & operations. Since launch, we’ve also done $12k in revenue.
Kollecto expenses (1st 6 months)
Here’s an update on how we’re building:
1. Sign up flow: Using Typeform
Typeform “logic jumps” let us can ask different questions based on the user’s previous answer.
.
2. Backend automation: Using Zapier, Pipedrive & Campaign Monitor
We use Zapier to automatically read the Typeform questionnaires & organize clients into lists in Campaign Monitor.
Zapier also reads the Typeform questionnaires & automatically organizes clients into our CRM software (Pipedrive)
Augmented Reality: Using ArtBeamer
ArtBeamer is an augmented reality app that lets clients see what art actually looks like on their walls. We pay $40/mo for a customized app for our clients.
.
An Email UX: Using Campaign Monitor
Our entire UX is via email!
And it works well – there are no logins, no passwords, easy analytics, and the experience feels natural (If you hired a personal assistant, you’d expect them to communicate with you via email. This isnt much different).
Clients get weekly personalized emails from their Advisor. They can request more info on any piece, give feedback, or review art from previous emails using buttons witin the email.
Startup explainer video: Using Videolean, Fiverr, & iMovie
A few months ago, I discovered Videolean.com– a site for DIY explainer videos. Videolean had some great templates, but more than anything, I liked the startup explainer video they used for their own business. So I contacted them & asked for it. 🙂 They sold it to me & customized my scenes for $200.
I rearranged scenes using iMovie, wrote a script, & paid $5 for a professional voiceover on Fiverr.
The video cost $205 total.
It’s not perfect, but I think it’s a good first video.
When I started Kollecto, I envisioned a tech-heavy platform for matching first-time collectors with an art advisor. I’m so glad I didn’t start writing code for that… because we’ve pivoted this service dozens of times.
Spending 0% of time on coding and 100% of time learning what my customers want has been really rewarding & successful.
