Localize.city is the US branch of Madlan, an Israeli AI-led company that provides insights about real estate and neighborhoods. I worked on a team to complete an extensive research and content audit project for Localize.city as part of the UX Writing Academy by UX Writing Hub.

Localize.city's staff UX writer asked us to focus on the experience of home-buyers, and work toward an overall goal of increasing user signups.

My work on this project included:



Competitor analysis

I took the lead on our competitor analysis. I created an airtable to track our findings and with the help of two teammates I analyzed experiences across four major competitors.

  • Redfin
  • Zillow
  • Trulia
  • StreetEasy
  • Discover industry trends
  • Understand how competitors approach and solve problems
  • Identify opportunities for Localize.city to stand out

Competitor anaylsis airtable

Competitor trends summary

Onboarding and signup flows
  • Only an email address is needed for registration. Passwords and other personal info are optional.
  • To use the mobile app there is a 1-3 step onboarding flow that can be skipped.
  • Profile setup is not required to use the apps or sites, but will help get more personalized search results.
404 pages
  • Some competitors display a standard 404 page or redirect users to the search page with an error message 'We could not find this area' depending on the type of bad link entered.
  • Competitor 404 pages were not always on-brand
  • Localize.city does not currently have any 404 page.
  • Language in competitor error messages is straighforward, but not especially helpful or conversational.
  • Contact forms are nearly identical across all competitors, but search forms vary widely.
  • Contact forms generally do not indicate required or optional fields.


  • 3 out of 4 of the top competitors are all operated by the same company - Zillow Trulia, and StreetEasy. Localize.city provides a new voice and perspective.
  • There is an opportunity to beat the competition by providing more useful and conversational error messages and CTAs.
  • Zillow, Trulia, and Redfin have a national audience. Only StreetEasy has a narrower focus on NY and NJ. StreetEasy provides some limited local insights mostly related to public transit. Localize.city can stand out as a hyper-local real estate advisor that operates from the renter/buyer perspective.

Conversation mining

Two teammates and I explored different spaces online and transcripts of user interviews to become familiar with the conversations users are having about real estate sites and the home buying process. We used airtable to organize our findings.

Conversation mining airtable


  • Users are looking for up-to-date and trustworthy information, tailored to their needs. They want to know things that agents can’t or won’t tell them.
  • Users are informal, direct, and vivid. They speak frankly about their feelings and anxieties. They are sometimes a bit sarcastic, and distrust agents’ jargon.
  • The main pain points we identified were being unsure of the site’s credibility, hidden drawbacks associated with properties, and concerns about safety and security.
  • Users can get very emotional talking about their homes. It’s something that touches them personally. This should be considered in the voice and tone guide.

Content audit

Five of us worked on an audit of Localize.city's content with a focus on the buyer's experience. We organized our review by user journeys and recorded our findings in an airtable.

Content audit airtable

My main area of contribution was in the mobile onboarding flow. I begn by mapping out the flow:

The primary user journey for a home buyer accessing the site via mobile for the first time.

The onboarding flow mapped with Whimsical

As I went through the flow I identified several pain points:

  • On the home screen, tapping "Buy" and the banner at the bottom advertising Robin, the AI assistant, direct the user to the same screen. Presenting both options suggests they will have different results, but there is no way to bypass Robin.
  • The welcome screen includes a secondary CTA "Skip." However, tapping it sends the user to the following screen.
  • The entire flow takes 8 screens to complete. This is much longer than competitor onboarding flows which are only 1-3 screens that are optional. Users may get fatigued and drop off.
  • In the final step, there is no obvious way for a user to opt-out of text messages. If users are reluctant to hand over personal details, this will lead to high drop-off rates.

Rewriting the onboarding flow

Localize.city's UX writer told us that they experienced high drop-off rates during their onboarding flow. I approached revising the flow to address the pain points above by reducing user fatigue and giving users more control over their personal data.

Before and after images of the first screen in the onboarding flow

Screen 1

Features: I've removed the banner for Robin, the AI assistant.

Since both the Buy button and Robin take the user to the same screen, the banner at the bottom of the screen is redundant.

It may also cause confusion if the user does not want to use Robin and taps 'Buy' which suggests a self-guided search in this context.

Before and after images of the second screen in the onboarding flow - welcome screen

Screen 2

Heading: "Greetings" is an archaism that sounds like an alien or robot. This may be intentional since Robin is an AI, but it would be worth testing a more familiar alternative like "Welcome."

Body: The revised version is friendlier and removes the second paragraph. It could erode trust if the expectation of a quick process is not met - this flow is at least twice as long as that on competitor sites.

CTA: "Skip" only sends the user to the following screen. "Browse homes" will skip the entire flow for users who are eager to begin their search right away. The word browse also suggests more casual use of the site. Serious buyers will still want to target their searches with Robin.

Before and after images of the third screen in the onboarding flow - neighborhood selection

Screen 3

Features: I added a progress tracker to every step so the user will know how much longer the flow will take.

Heading: The revised message makes it clear that more than one neighborhood can be selected.

CTA: I included a secondary CTA "Skip" so that users know they can move on without selecting a neighborhood. I added this to all the steps except the last one.

Before and after images of the fourth screen in the onboarding flow - basic filters

Screen 4

Heading: “Nice choice!” seems overly enthusiastic and won’t work well if a user chooses multiple neighborhoods in the previous step.

Features: The price range slider looks great, but is difficult to use on smaller screens, especially if the price range is on the smaller side. I added fields so that the user can manually enter dollar amounts if they have trouble with the slider.

Before and after images of the third screen in the onboarding flow - topics of interest

Screen 5

Heading: The original heading asks the user what they are interested in. A user may not feel like they're interested in any of these options. I've reframed these as "special requests" which is more in line with how users are likely to feel about this step.

Microcopy: I expanded the microcopy here to better convey the value in making selections in this step.

"Don't worry" is a phrase I typically avoid. Users don't like being told how they should feel. Often, when a person is told not to worry they start to think they should worry.

Before and after images of the fourth screen in the onboarding flow - commute details

Screen 6

Features: I combined the 6th and 7th screens of the original flow.

The original address entry screen didn’t visually fit in with the rest of the flow, which can be jarring.

Heading:I added an additional heading which eliminated awkward phrasing in the original version - “commute address” is not a common expression.

Before and after images of the third screen in the onboarding flow - signup

Screen 7

Heading: The revised heading is user-focused. It’s not about what Robin is going to do, it’s about what the user is doing.

Body: I've made it more clear that this is an optional signup form. This is not indicated in the original version.

Users may not want to receive texts and there’s no way to opt out of them in the original version. My version puts the user back in control of their personal data.

A/B testing

Localize.city's main KPI is user signups. To increase user adoption we offered suggestions for testing the primary account registration screen on desktop.

Image of original version of registration screen

Original screen

image of first option for A/B testing

Option A

Option A:

Mood: The user is non-committal.

Hypothesis: The user is not ready to give out their phone number.

Solution: Remove the phone number field and allow the user to sign up with just their email address.

Reasoning: Sharing an email address is less of a committment than sharing a phone number.

image of second option for A/B testing

Option B

Option B:

Mood: The user is apprehensive.

Hypothesis: The user needs more context before they will decide to share their personal info.

Solution: "Get personalized email or text recommendations and never miss the latest listings"
Mark the phone number field as optional.

Reasoning: The suggested copy gives context by explaining the value of signing up. Giving the option to omit the phone number also builds trust and puts the user in control of their data.


Localize.city has a real opportunity to rise above the competition by providing trustworthy and unique insights into properties that aren't available elsewhere. We hope to have improved the user experience, especially during registration and onboarding, so that they see higher rates of user adoption.

Get in touch

Interested in working with me on your next project? I'd love to hear from you.

Email me