Lula Weekly Status, Weeks of July 27 and August 3, 2026

From: Gemma Period: 2026-07-27 to 2026-08-09

This update covers two weeks. The last status went out for the week of July 20, so everything since then is here.

Executive Summary

The largest piece of work in these two weeks is the homepage. Twenty-three sites were torn down and read against their real organic performance, and out of that came a full rebuild recommendation: an anchor sentence, a new hero, fourteen blocks with the copy written to be used, four things to cut, and a sequence so it does not have to ship at once. The document is honest about its own limit, which is that it converts traffic rather than creating it.

Alongside it, the Kansas City edition of the maintenance cost index was built as a working example rather than another description of the idea. It runs entirely on public primary sources, so it can ship before Lula's own data export exists, and it corrects seven claims currently circulating about the KC market.

The after-hours answering service landing page was built as the conversion destination for the paid push with dayparted bids. The reason it is worth building rather than pointing the spend at an existing page is that every page-one result for the commercial term is a how-to telling property managers how to handle after-hours calls themselves. Nothing on that page positions a service that does it for them, which leaves the commercial slot open.

The technician scheduling article was rebuilt and republished. Last update reported the rewrite as a workforce and burnout asset. On a closer read of the results page, the keyword the page is named for is a software-evaluation search, so the piece was rebuilt again to match what the page actually has to answer. It now leads with a direct definition, adds a "what to look for" evaluation section, and keeps the burnout narrative as supporting argument. A 2021 turnover statistic it leaned on was replaced with 2025 and 2026 figures, since the claim it supported does not hold on current data.

The /services page came off the bench and is finished: the full paste-ready build, the CSS moved somewhere the classic editor cannot corrupt it, the schema layer, and the coordination block rewritten with the tier names out. It needs someone to paste it, not more work from me. Alongside it, the two competitor reads you asked for both came back with the same shape of answer, which is that neither competitor is as strong as their numbers look, and one of them has left an answer-engine opening wide enough to walk through.

Two research pieces changed direction rather than producing content, and both saved work. The property maintenance tracking software outline was stopped before drafting, because four of its five proposed sections duplicate pages that are already live and its keyword set collides with the push already underway on the software comparison page. The extreme weather set went the other way: the heat article moved from a spring 2027 publish to January 2027 once the LA County indoor temperature mandate turned up in research with an enforcement date attached.

Measurement is mixed and honest about it. Referring domains are the standout, up 77 in two weeks to 1,184. Search Console clicks eased about 6 percent on higher impressions, and average position slipped from 6.8 to 7.5, which is the number to watch rather than the clicks.

Wins This Period

Homepage Rebuild Recommendation, Off 23 Site Teardowns

Twenty-three homepages torn down and read against their actual organic performance, across property maintenance, proptech, field service, marketplaces, fintech, and B2B SaaS. Delivered as three pieces: the teardown itself, a screenshot gallery of the specific patterns worth copying, and a separate read on strong sites outside the category. The recommendation and a design mockup sit on top of them.
The problem the page has to solve is a portfolio problem. The homepage is carrying Coordination in both its tiers, the Pro Network, Flat Rate, Make-Readies, LuMi, Foresight, and the service provider side. Breadth pushes copy up the ladder of abstraction until it says nothing, which is how a homepage ends up at "maintenance made easy."
The fix is a single anchor sentence everything else hangs off: every maintenance job, from the resident's first message to the closed invoice, run by one team. That sentence is what makes Coordination, the Pro Network, Make-Readies, and LuMi read as one system rather than four line items, and it sets the page's boundary. Anything that sells software to a team the operator already employs sits outside the main flow and gets its own module.
Recommended hero: "The maintenance department you don't have to hire." The current claim is available to every competitor and most of them are already making it. This one is a category claim rather than an adjective, it keeps the headcount argument that actually unlocks budget, and it names the decision a property manager is really making, which is whether to hire another coordinator. Lula is the only company in this space selling the triage, the technicians, and the price certainty as one thing.
Fourteen blocks specified in page order with the copy written to be used, each naming the site the pattern came from. The ones worth calling out: a live work-order counter above the headline, a second call to action offering a free maintenance assessment so the researching buyer has a door that is not a sales call, replacing the cartoon illustration with the coordination product carrying real state, a pro profile card that proves the word "vetted" instead of asserting it, the comparison table filled with numbers on both sides, a make-ready module, a Foresight band placed as a deliberate turn in the page for a different buyer, and publishing the Flat Rate catalog.
The cheapest high-value change is one sentence. Adding a methodology line under the stats already on the page converts three marketing claims into citable statistics, which is what gets them repeated by a trade publication or pulled into an AI answer. No design, no engineering.
Flat Rate is the recommendation I would argue hardest for. It is currently invisible above the fold and absent from navigation, which is a strange place to keep the one thing no competitor can match. Publishing even a representative set of repair prices proves the claim rather than asserting it, and concrete repair costs are exactly what property managers and answer engines search for. Nobody in the category publishes them.
Four things recommended for removal: the cartoon illustration, generic benefit copy (if a line survives being moved to a competitor's site unchanged, it is not working), the service provider path in the body copy, and any unsourced round number.
Sequenced so it does not have to ship at once: the methodology lines and the comparison table data this week, the headline and product architecture next sprint, the hero visual, pricing page, and live counter at quarter scale.
One caution stated plainly in the document. This rebuild converts the traffic Lula already has considerably better. It will not create traffic. At around 2,361 organic visits a month across 161 keywords, the homepage is not the constraint on growth, the absence of a page system is. Two of the benchmark sites convert 70 and 55 visits per referring domain against Lula's 2.1, and they get there through systematic pages rather than better hero copy. This is the conversion half. The acquisition half is the make-ready cluster, the cost index, and whatever trade or location page system follows them.
Open items before any of it publishes: the one-line difference between the two coordination tiers, the scope of the Flat Rate catalog, and the placeholder figures in the comparison table and the named-outcomes block, which show shape only and must be replaced with verified Lula data.

Kansas City Maintenance Cost Index Built as a Working Example

The first worked metro edition of the cost index, built to show what the program produces rather than describe it. Kansas City MO-KS, 14 counties.
Built entirely on public primary sources, which is the point: the American Community Survey, NOAA climate normals, the Bureau of Labor Statistics, Census County Business Patterns, HUD's Kansas City market analysis, and the municipal code itself. It can ship before Lula's own export exists, and Lula's data makes the next edition stronger rather than being the thing that unblocks it.
Five findings, each one a correction to something the market currently believes. The single-family rental stock is 67.1 percent pre-1980 against 60.3 percent nationally, even though metro-wide rental housing reads slightly younger than the national median, because blending garden apartments with prewar houses washes out the signal. Homes heat with utility gas at 65.3 percent against 46.5 percent nationally. The metro averages 10.9 hail days a year, which exceeds NOAA's own hail alley. Roofers are the only under-supplied trade at a location quotient of 0.92, which sits in the market with the highest hail exposure in the report. And the Missouri and Kansas sides are a regulatory mirror image of each other, with federal lead rules changing administrator at State Line Road.
Seven claims currently circulating in this market are contradicted in the report, three of them established by full-text search of the current municipal code. The widely repeated 79 freeze-thaw cycles figure attributed to NIST is untraceable, with no publication, date, or methodology locatable, so the report derives its own number from NOAA daily observations and says so.
The finding I think is genuinely unpublished anywhere: the metro with an unusually high concentration of combustion appliances has no ordinance requiring carbon monoxide alarms in long-term rental housing. The city gives the detectors away free through its Healthy Homes program instead of mandating them.
The operational read is the part a property manager will use. A portfolio concentrated in Jackson or Wyandotte County carries roughly three quarters pre-1980 construction, meaning original supply lines, undersized panels, and lead paint exposure. The same portfolio in Olathe or Platte County is under half. Two books of business in the same metro, an hour apart, carry materially different maintenance profiles. Roughly 76,500 Kansas City single-family rentals were built before 1980.
Every figure is sourced, dated, and free to cite, with the precision caveats kept in rather than smoothed over. That is the whole design: this asset is graded on whether it gets cited, so the sourcing is the product.

After-Hours Answering Service Landing Page Built

Written as the destination for the paid after-hours effort. With dayparted bids, a meaningful share of visitors will land on this page at 10pm or later while they are personally on the hook for the call that just came in. The hero is written for that person in that moment rather than for a daytime software evaluator.
The commercial slot is open. Primary keyword is after hours answering service property management at 250 a month and a keyword difficulty of 1, with property management answering service at 1,600 in the same space. Every page-one result today is a how-to guide. A page that answers the same question with the service as the answer is doing a different job than the guides it sits next to.
The three-asset overlap is resolved on paper and needs one edit to be resolved in practice. Three assets sit in this space: this commercial page, the after-hours maintenance calls article, and the answering service evaluation article. Each now has one intent assigned. The article draft currently targets the commercial term, so it gets re-pointed at the how-to term before either goes live, and it links up to the new page.
Delivered as copy and as a paste-ready build, with three headline variants for Tyler to test.

Technician Scheduling Article Rewritten and Republished

Republished on 2026-08-07. This is the one scheduling spoke Google had crawled and never indexed.
Rebuilt to match search intent. The page had been written as a burnout and staffing argument while the keyword it is named for is a software-evaluation search. The rewrite leads with a direct definition of technician scheduling software above the fold, adds a section on what the software does, and adds a "what to look for" evaluation section using category framing with no competitors named. The property manager audience and the Foresight positioning carry through.
Two FAQs added covering ground the body does not: how scheduling differs from route optimization, and how the software works alongside an existing PMS.
A statistic was corrected rather than kept. The maintenance turnover figure the piece leaned on measured 2021, and the claim it supported does not hold on current data. Both were replaced using the 2025 RCLCO compensation survey published by NAA and NAA's Q2 2026 apartment labor report.
Indexing requested in Search Console the same day. The first thing to check is whether the page leaves crawled-not-indexed at all, since that gates everything else. Full measurement window opens 2026-09-06.

Extreme Weather Set Drafted, and the Heat Piece Moved Up Nine Months

Freeze-event article drafted, covering what happens to a portfolio when a freeze hits every unit on the same night: the call flood and triage problem, why the local vendor market saturates at exactly that moment, and why markets that freeze rarely take more damage per event than cold-climate markets. Built on verified 2025 and 2026 claims data from State Farm and Verisk rather than a seasonal checklist, since checklist ground is held by the Red Cross and Consumer Reports and is not winnable. Targeted at a September publish to land inside the freeze-prep window.
Graded on citations rather than sessions. No rental-framed keyword in this space carries search volume, so this is an answer-engine play and the measurement is set accordingly up front.
Heat article drafted, and research changed both its shape and its date. Heat is a duration story rather than the freeze piece with the temperature inverted: cooling systems run at maximum output for weeks, failures accumulate across a season, the power grid becomes part of the maintenance risk, and cooling is starting to be written into habitability law. LA County passed an indoor temperature mandate in August 2025 requiring rental units to be kept at 82 degrees or cooler, with enforcement beginning 2027-01-01 and a carve-out for owners of ten or fewer units until 2032. That deadline moved the publish from April or May 2027 to January 2027, so it lands while property managers are working the problem.
What is deliberately not in the draft: Maryland and California cooling requirements were cut because they could not be verified on source. They would strengthen the piece if Lula can confirm them.

Maintenance Cost Index Proposal Passed to Nate

The full proposal and an interactive concept went to Nate on 2026-08-04, following the Slack agreement on 2026-08-03 to pursue a quarterly automated export of aggregate work-order data rather than a live dashboard.
The thesis is a gap in the category. Every existing maintenance cost source is either survey-based or consumer retail pricing. Lula has real invoiced work-order data and nobody publishes it.
The flat-rate tension is resolved, and it resolves in Lula's favor. The index works as the evidence behind flat rate. The pricing page already claims that pricing stays predictable even when regional labor and material costs vary, with nothing published behind it. Metro variance is that proof. The firewall rule is that the index reports the market cost of the work and never Lula's charged prices.
Specified to the field level. Seven tables, with small-sample suppression and outlier trimming enforced in the query rather than left to editorial judgment. Twenty-four URLs, not the 750 trade-by-metro combinations, which live as filter states rather than pages.
Measured on citations and referring domains. Sessions are the wrong grade for this one: the statistics queries here run 30 to 70 searches a month. The comparison that matters is that a competitor's 2017 renters report still holds 326 referring domains while Lula's best content page holds 38.

Two Research Calls That Prevented Work

The property maintenance tracking software outline was stopped before drafting. The brief's premise was that the library lacks a piece on tracking, dashboards, and the move away from manual methods. All three are live pages already, and four of the five proposed sections duplicate them. The proposed keywords all resolve to the same parent topic as property maintenance software, which is the exact term the software comparison page is being pushed from position 11 to page one on, so a second page would work against that push. The results page also wants product and solution pages rather than blog posts: five of the top eight are product pages, and the stronger of the two articles sits at position 9. Recommendation is to fold the tracking query into the existing hub, with a net-new article on a genuinely different angle available if Nate wants net-new content this cycle.
One framing was rejected on ICP grounds. The outline's portfolio-size framework, spreadsheets under 20 units and automation past 50, targets an audience Lula does not sell to and would pull in the freebie-hunter segment the source-of-truth doc flags.

Full Blog Library Inventoried for a Strategy Relook

Gemma pulled the complete blog inventory manually to re-examine strategy against what is actually published rather than against what the roadmap assumes. It covers roughly 70 live articles with URL, headline, last-updated date, internal and external links, primary and secondary keywords, search intent, and the source document for each.
Why it was worth doing by hand. Two of the calls this period, the tracking software stop and the technician scheduling intent correction, both came from checking a proposed piece against what is already live. The inventory makes that check fast and repeatable instead of a per-brief investigation, and it surfaces the stale-date and thin-metadata cases in one view.
First read from it: a meaningful share of the older announcement and partnership articles carry no keyword or intent assignment at all, and several 2025 pages have not been touched since publish. Both are candidates for the refresh queue rather than new drafts.

The /services Page Finished and Handed Over Ready to Paste

Full paste-ready build delivered, hero through FAQ, for the consolidated Property Maintenance page. Leads on "Property Maintenance, Fully Coordinated" with an animated work-order lifecycle card running submitted, triage, scheduled, resolved.
The page CSS was moved out of the page content and into Additional CSS. The classic editor rewrites newlines inside style tags into line-break tags, which silently invalidates the CSS. Additional CSS is never touched by the editor. Delivered both ways, as a separate stylesheet and as a minified single-line style block, so it can go in whichever route suits the person doing the paste.
Schema layer written, adding only what Yoast cannot generate and redefining nothing it already outputs: the coordination Service itself, an OfferCatalog of the 29 named trades and services on the page so answer engines can match "does Lula do X" queries against a machine-readable list, areaServed across the 56 markets, and the five visible FAQ entries verbatim. It references the existing Yoast entities rather than duplicating them, and was verified against the live page and both draft versions, so it is safe to paste on any of the three.
Coordination block rewritten with the tier names removed as separate named offerings, per the page guardrail, reframing the two feature sets as one coordination system with coverage that scales.

Two Competitor Reads You Asked For

The first read: an AI-native PMS whose search footprint is much smaller than the brand implies. Roughly 1,200 organic visits a month, more than half of it people typing the brand name or a homophone. Strip that and it is 400 to 500 qualified non-brand visits a month across a 282-page site.
The finding worth acting on is the ratio. They hold about 250 Google AI citations against four real non-brand organic rankings. That is not a deliberate answer-engine program. It is what happens when you publish a lot of moderately good definitional content in a category where Google's AI surfaces are short of sources. The same opening is available to Lula, with better material behind it.
They are also blocking most AI crawlers in robots.txt and appear not to know it. A default managed block is disallowing the major AI crawlers. Their Gemini citations sit at exactly zero and ChatGPT at nine, while Google's own surfaces, which the block does not touch, sit at 250. Worth a check that nothing similar is running on lula.life.
The second read: a large competitor whose numbers turn out to be an accounting artifact. Of about 5,092 US organic visits a month, roughly 4,284 (84 percent) are people searching the brand of a company they acquired in 2023, and another 606 are their own brand. Genuine non-brand category traffic is about 32 visits a month, and half of that lands on a business intelligence post unrelated to property maintenance.
The "higher value" number does not survive contact either. Their $30.7K a month traffic value is 76 percent one page, a merger announcement ranking first for an acquired brand term carrying a $12.90 cost per click because competitors bid against it. They do not pay that cost per click and are not converting that traffic. Strip two pages and 92 percent of the traffic value disappears. Their trend is down 15 percent year over year with no month above the prior year's baseline.
On answer engines Lula is ahead of them on every surface but one: 50 AI Overview keyword citations across 13 pages against their 10 across 8, and 21 AI Mode citations against their zero. Their bigger footprint is in the channel that is shrinking.
One idea came straight out of the first read and is now a card in Trello. Neither competitor has a subject-matter expert, a byline, or any first-party expertise anywhere in their content. A recurring Pro interview series would put exactly that on the site: a short question set to a different Lula Pro each month, cut into blog, social, YouTube, and newsletter. Sourced from the people who actually complete the work, which is the one thing a software competitor cannot copy.

Measurement Highlights

Rankings and AI Visibility

The software comparison page is picking terms back up. property maintenance system entered at position 8 (volume 150) and property maintenance software programs returned at position 7 (volume 100). Both had dropped out of the tracked range in the last update, so this is a recovery on the page that has been moving most week to week.
The same page lost an AI Overview citation on property maintenance app (volume 250) while improving from position 12 to 11 organically. This is the presence-share pattern reported for the last three weeks: rankings hold or improve while citation presence toggles. It is now consistent enough to treat as the normal state of these results rather than as an incident.
Rental inspection checklist gains held and extended. rental walkthrough checklist pdf moved from 28 to 22 (volume 250), and quarterly rental inspection checklist pdf returned at position 10 (volume 100). That page remains on the watch list from the sharp drop after its June edit, so these are the first clear positive signals on it.
Make-ready is split. make ready service returned at position 4 and apartment make ready services near me improved from 6 to 5 (volume 100). Against that, make ready services near me (volume 100) sits at 20, down from 19 last week and down from position 8 two weeks ago. One "near me" query improving while its close sibling falls twelve places points at results-page volatility on local intent rather than a page problem, and the 16 city pages are the right structural answer either way. Worth one more week before acting.

Search Console (last 7 days, 2026-07-29 to 2026-08-04)

Clicks 564, down about 6 percent week over week, on impressions up about 4 percent to 26,136. Click-through eased from 2.39 to 2.16 percent, and average position moved from 6.8 to 7.5.
Average position is the number to watch here, ahead of the click dip. More impressions at a lower average position is the signature of ranking for a wider set of weaker queries, which is normal while a library expands and is only a problem if it continues. This is the first week it has shown, so it is flagged rather than diagnosed.

Organic Traffic

Estimated organic traffic is about 2,296 a month, easing from roughly 2,345, with 88 keywords in the top 3, down from about 93, across 150 tracked organic keywords.
The movement is small and sits inside the same three pages diagnosed on the July 20 scan. Nothing new has broken.

Backlinks

Referring domains reached 1,184, up 77 across the two weeks (1,107 on July 20, 1,153 on July 27, 1,184 on August 3). This is the strongest and steadiest line in the report: 844 on June 1 to 1,184 now, a 40 percent gain in nine weeks, and unlike earlier months the gains are arriving weekly rather than in bursts.

Now in Your Court

Homepage rebuild: the decision to make is how much of it you want and in what order, since it is sequenced to ship in pieces. Three things are needed from you regardless: the one-line difference between the two coordination tiers, the scope of the Flat Rate catalog, and real figures for the comparison table and the named outcomes, since the ones in the document show shape only. The two changes I would take this week either way are the methodology lines under the existing stats and the real numbers in the comparison table. Neither needs design or engineering.
Publishing the Flat Rate catalog is a separate call and a bigger one. It is the recommendation with the most upside and the one most likely to get pushback internally.
Kansas City cost index: review, and tell me whether this is the format you want the program to take. If it is, the next question is which metros follow and in what order.
After-hours landing page: review with Josh, and confirm the intent split across the three after-hours assets so the article draft can be re-pointed before either publishes.
Maintenance Cost Index: the six items in section 8 of the proposal, the most important being contract permission for aggregate de-identified data use and who owns the data model. On a yes to the contract question, section 5 becomes an engineering ticket and I draft the hub structure so the first export has somewhere to land.
Extreme weather set: review both drafts. The heat piece needs a specific legal read on the compliance section. The freeze piece would improve materially with work-order volume data from a past freeze event, and the Maryland and California cooling requirements need confirming or they stay out.
/services page: everything is delivered and ready to paste. What is left on your side is the three small fact confirms, the pre-redirect citation check on the two pages being consolidated, then publish and set the redirects.
Tracking software research: decide whether to fold the query into the existing hub or commission the net-new angle.
AI crawler check on lula.life: worth confirming no default managed block is disallowing the major AI crawlers, since that is exactly what one competitor did without noticing.
Pro interview series: the card is in Trello. It needs a yes or no before I scope it.
Water heater and plumbing spokes: still waiting on the cost figures for both, or those paragraphs get cut. No invented numbers either way.
Preventive maintenance checklist: still needs review, and the 80 percent first-trip claim resolved before it can publish, since source-of-truth records that as an industry benchmark rather than a verified Lula proof point.
Still open from earlier updates: the recognition and announcement articles awaiting facts and quote sign-offs, the services consolidation confirms and the citation check before its redirects, the Flat Rate rework plus the HVAC page claims, the Cluster 2 lead magnet approval and placement, the reviews page, and the go-ahead on the PMS integration question-and-answer blocks.

Coming Up

Take the homepage recommendation into whatever shape you want to act on: the two same-week changes are ready to hand to whoever edits the page.
If the Kansas City edition lands well, scope the next metro and settle whether these run ahead of Lula's data export or wait for it.
Watch whether the technician scheduling page leaves crawled-not-indexed after the re-crawl request. That is the gate on everything else for that page.
Work the blog inventory into a refresh queue: the 2025 pages with no keyword or intent assignment, and the pages untouched since publish, ranked against the Q3 priorities.
Ship the after-hours article re-point and the internal link up to the new commercial page.
Publish the reactive-maintenance spokes and the preventive checklist as the open data points get resolved.
Give the average-position slip one more week of data before deciding whether it needs a diagnosis.