Haferman 2026 Year-to-Date Review
Haferman Water Conditioning, Inc.
Media Relations Agency
Reporting Period
Jan 1 – Sep 30, 2026
Year-to-Date Review · January through September 2026

2026 so far — leads held level while the path to them moved from the blog to the phone and the map

Through September, Haferman's website produced 860 leads, 372 of them new-business requests, against 841 and 364 over the same nine months of 2025. Underneath that flat line, the year had three distinct chapters: a strong first quarter with organic search at a peak, a spring and summer in which Google began answering Haferman's how-to articles on its own results page and blog traffic fell by two thirds, and a September in which the Google Ads campaign drifted into display placements and paid calls halved. Through all of it the phone outperformed the website form: 3,029 tracked phone contacts (1,626 from website visitors, 1,403 from the Google Business Profile) against 860 forms, about three and a half to one. This review sets the full-year context for the 90-day plan.

The year at a glance

Tracked Phone Contacts
3,029
website calls 1,626 · profile calls 1,403
Website Leads
860
▲2% vs Jan–Sep 2025 (841)
New-Business Leads
372
▲2% vs 2025 (364)
Paid Cost per Contact
$14
$23 per first-time contact · $13,261 spent
Organic Search Clicks
820
▼63% from the March peak (2,206)
Profile Direction Requests
1,890
▲Sep 225 vs Jan 183
Google Reviews
4.9 ★
247 new this year (243 five-star) · 1,631 total
Facebook Views
770,521
122 posts · $2,007 boosted · +33 followers

The year in three chapters

January – March: the high-water mark

Organic search peaked. Search Console clicks ran 2,077, 2,073 and 2,206 a month, the best three months in the data, with an average position of 8 in March. Nearly all of it was how-to traffic: standing water in the salt tank, low pressure, salt bridges.

Leads were strong. 108, 124 and 89 website leads; February was inflated by a bot-spam episode (46 leads with no channel). Paid calls ran 98–108 a month at about $1,356 of spend.

Google Ads conversions looked their best (163 in January) before the ServiceTitan import stopped reporting in March; the drop that followed in the Ads dashboard is partly a measurement change, not a lead change.

April – August: the blog loses its audience

Search clicks fell every month from 2,206 in March to 907 in August, and average position slid from 8 to 18. Organic visits to the how-to articles went from 1,727 in January to 635 in August, down 64%. Google's AI Overviews now answer those questions in place; Haferman is cited in 20 of them but no longer gets the click.

The pages that sell did not fall. Organic visits to the home, product, service and contact pages held between 459 and 674 a month all year. Organic callers held too: 107 in April, 85 in August.

Fair Season (July–September, $1,810) put Haferman in front of 561,380 people on Facebook and produced six QR-code water-test leads.

September: the paid calls halve

The week of August 31 the Performance Max campaign began serving display and YouTube placements. Impressions tripled to 30,065, clicks rose, and calls from paid visitors fell from 102 to 50 at the same budget. Cost per paid contact went from $13 to $26. A smaller version of the same drift happened in April (23,971 impressions) and cost less; September's was larger and landed on the busiest season.

Restoring the search placements is the first item in the 90-day plan.

All year: the map and the phone carried the load

The Business Profile grew while the blog shrank: 1,403 calls, 1,890 direction requests and 1,566 website clicks year to date, with September the best month for calls (212) and profile views (5,352).

Phone contacts ran 3.5 to one over forms. Call tracking recorded 1,626 website-driven calls, 1,048 of them first-time callers. Earlier briefs counted only the forms; that undercounted the year.

Reviews kept pace: 247 new reviews, 243 of them five-star, and half of them name the technician who came.

Organic search — the decline is the blog, not the business

Organic visits by page type, monthly

2,0001,5001,00050001,727620586674JanFebMarAprMayJunJulAugSep
Articles and FAQs (how-to)Home, product, service, contactService-area pages
Google Analytics organic sessions by landing page, grouped by what the page is for. How-to pages fell 65% January to September; the selling pages rose 15%. The 164 service-area pages drew 525 organic visits all year, which is the gap the rebuild's county pages address.

Search Console clicks, monthly

2,4001,8001,20060002,077820JanFebMarAprMayJunJulAugSep
Clicks from Google search results. 13,658 year to date. The slide starts in April and matches Google's rollout of AI Overviews on troubleshooting queries; Haferman's articles are cited in 20 overviews (SEMrush) but the answer is read on Google's page. Average position: 8 in March, 18 in August, 14 in September.

The pages people still arrive on organically, Jan–Sep

/3,025
/standing-water-in-water-softener-salt-tank2,238
/4-reasons-your-water-softener-may-be-causing-low-…1,152
/water-softener-salt-bridge906
/how-often-should-my-water-softener-regenerate604
/faq/q-how-do-i-manually-regenerate-my-kinetico-wa…567
/faq/what-should-i-expect-right-after-my-reverse-o…521
/softener-salt-vs-sidewalk-salt-whats-the-differen…334
/hard-water-problems-tips-for-limescale-removal318

One selling page (the home page) and eight troubleshooting articles. The articles still earn visits; they earn fewer each month and almost no leads. The plan keeps them, restructured as question-and-answer pages the AI engines can cite with Haferman's name attached, and puts the new effort into pages that convert.

What we track today is too narrow to judge this

Position Tracking covers 171 keywords, most of them troubleshooting phrases chosen in 2025 for the blog. It sees Haferman cited on repair questions and sees Culligan cited on buying questions, because those are the terms on the list. It does not see the local buying searches that matter most now ("water softener installation Lakeville," "who services Kinetico near me"), and it cannot see what ChatGPT, Gemini, Perplexity or Google AI Mode answer when someone asks who to call.

Two changes, both in the 90-day plan. Expand the Position Tracking list to the buying and local terms by county, so share of voice is measured on the searches the rebuild targets. And run the two SEMrush AI tests (AI Visibility Tracking, AI PR) for two months, so the November and December briefs show whether Haferman is recommended on buying questions and which sources the assistants draw on.

The goal for the rebuild: be cited on the buying questions the way Haferman already is on the repair questions. Question-and-answer formatting, structured data and the county pages are built for that.

Share of voice against Culligan — strong until May, then the board changed

SEMrush visibility, Haferman vs. Culligan, May 2025 – Oct 2026

SEMrush visibility chart: Haferman around 30% through May 2026, falling to the mid-teens by October; Culligan rising from the high teens to 44% from May 2026
Visibility is SEMrush's share-of-voice estimate across the 171 keywords we track (Haferman 23.7%, Culligan 44.8%, Morton 5.4%, WaterSmart 4.0% on Oct 13). Haferman led Culligan for the first twelve months of tracking, holding about 30% while Culligan sat near 15–20%. The lines cross in late May 2026, when Google began answering more of these questions itself and SEMrush began scoring who gets quoted in those answers. Culligan's national content library gets quoted more often; Haferman's rankings on the terms it already held did not move.

Haferman's content is showing up in Google's AI answers

Google AI Overview for 'can water softener cause low water pressure' citing Haferman Water alongside Carter Water, Parker & Sons, Aqua Clear, Quality Water Treatment, Benjamin Franklin Plumbing and Reddit
"Can a water softener cause low water pressure," October 13. Haferman's article is one of the cited sources and appears in the source panel next to Reddit and YouTube. The same is true for the salt-tank, salt-bridge and troubleshooting articles. These citations are where the lost blog clicks went: the answer is read on Google's page with Haferman's name on it, and the visit happens only when the reader wants more.

Website leads — level with 2025, with a different mix

All website leads, 2026 vs 2025

14010570350108124898585878811381JanFebMarAprMayJunJulAugSep
20252026
Form submissions on hafermanwater.com, all types. 860 in 2026 against 841 in 2025. February 2026 includes 46 bot submissions since blocked; August includes a buyout-calculator spike.

New-business leads, 2026 vs 2025

705235180483638413536346341JanFebMarAprMayJunJulAugSep
20252026
Water analysis and estimate requests, buyout-calculator submissions, fair tests and referrals. 372 in 2026, 364 in 2025. The other 488 leads this year were service calls (263), filter and supply orders (147), salt delivery and billing: current customers using the site as a service desk.

Where the 2026 leads came from

Organic search
352 · 40%
Direct / typed
221 · 25%
Google Ads
168 · 19%
Referral
41 · 4%
Unknown
78 · 9%

New-business leads split the same way: organic 162, direct 124, paid 63, referral 20. Direct is mostly people who already know the name; it includes radio, print and word of mouth, which is why tracked numbers on those channels are in the plan.

What people asked for, Jan–Sep

Water Analysis & Estimate297
Service Call263
Order Filters & Supplies147
Equipment Buyout68
Billing Questions39
Salt Delivery39
(not stated)7

Only four website visits all year came referred from kinetico.com. Kinetico's referred leads arrive through its dealer system, not the website, so the decline Kinetico reported is not visible here; we still need those monthly counts from January 2025 to size it.

Phone leads — the channel the forms never showed

Tracked website calls by visitor source, monthly

12090603001025074JanFebMarAprMayJunJulAugSep
Google Ads visitorsOrganic search visitors
CallRail calls placed by website visitors, by how they arrived. Direct and other sources add 4–13 a month. 1,626 calls year to date, 1,048 first-time callers. Paid calls averaged 95 a month from January through August; September's 50 is the placement drift.

Google Business Profile: calls and direction requests, monthly

260195130650175212225JanFebMarAprMayJunJulAugSep
Calls from the profileDirection requests
Calls placed from the Maps or search listing without a website visit. 1,403 calls and 1,890 direction requests year to date, plus 1,566 clicks through to the site. The top searches that surface the profile: water softener (718), kinetico water softener (639), water softener system, water softener salt, reverse osmosis system.

What the full year says about the lead count

Counting forms alone, 2026 looks flat. Counting every tracked contact, the year produced 3,889 leads and service contacts: 860 forms, 1,626 website calls and 1,403 profile calls. Those calls include current customers booking service, which is why the ServiceTitan connection in the plan matters: it is the only way to say how many of them became jobs. Until then, the two numbers that stay on the first page of every brief are first-time phone contacts and new-business forms.

Advertising — steady spend, one expensive month

Google Ads spend, monthly

1,8001,35090045001,3561,650JanFebMarAprMayJunJulAugSep
$13,261 year to date across Performance Max (11,452), the Apple Valley Performance Max test (910), a managed Minnesota campaign (758) and remarketing (141). Spend rose from $1,356 to $1,650 a month over the year.

Cost per paid contact (calls + forms), monthly

302215801226JanFebMarAprMayJunJulAugSep
Google Ads spend divided by tracked calls and form leads from paid visitors. $14 for the year, $23 per first-time contact. It held between $10 and $14 through August and jumped to $26 in September. Google's own conversion count (163 in January, 30 in September) is not used here because the actions are split, two are misconfigured, and CallRail's calls are not yet feeding Google.

Facebook and Instagram

122 posts, 770,521 content views, 33 new followers (554 total). Views come in bursts when posts are boosted: $2,007 in boosts this year, almost all of it the Fair Season campaign in July–September, which reached 561,380 people. Unboosted posts average about 40 views.

The best-read posts of the year were benefit-led: whole-home water treatment explained (67,739 views, boosted), PFAS in Apple Valley (16,996), the Chaska wellness expo (16,792), and "a real person answers" (8,913). That pattern is what the always-on campaign in the plan buys every month instead of in bursts.

Instagram reached 18,929 accounts in April when a post was boosted and under 1,000 in most other months, with 5 website clicks all year.

Reviews, and what they say

247 new Google reviews this year (243 five-star; 238 in the same months of 2025), 1,631 in total at 4.9. Replies: 72% of this year's reviews answered, 97% in the last 90 days.

Half of the reviews since 2023 name the technician who came; a third say friendly; a quarter say he explained the system or answered every question. Almost none mention the product by feature. That is the brand, and the messaging guide is built from it.

By month: Jan 27 · Feb 23 · Mar 32 · Apr 26 · May 28 · Jun 27 · Jul 36 · Aug 27 · Sep 21.

What the year tells us about the plan

1

The organic loss is a content-type loss, so the fix is pages, not more posts

How-to traffic went to AI answers; the selling pages and the local pages held or grew. The rebuild's county pages, booking page and restructured product pages are aimed at the buying searches Haferman does not rank for today.

2

Paid search worked at $12–$14 a contact for eight months; September was a placement problem

Restricting Performance Max to search, adding a Search campaign on the terms that surface the profile, and feeding CallRail's calls into Google are the no-cost items that put paid calls back in the 90–100 range.

3

The Business Profile is the channel that grew all year without a budget

1,403 calls and 1,890 direction requests with no booking link and no posting schedule. Both are in the plan at no cost; the Local Services Ads test puts a paid call button above the same map.

4

The phone is the lead channel, so measurement has to follow it

Forms were 22% of contacts this year. Cost per lead, the November and December targets, and the December budget decision are all stated in calls and first-time contacts; the ServiceTitan connection adds booked jobs behind them.

What this review can and can't compare

Monthly trends run January through September 2026 for every source. Year-over-year comparison exists for website leads (the sheet goes back to April 2024) and for September 2025 onward on everything else; the Google Ads, Analytics, Search Console, Business Profile and Facebook feeds were connected with history from September 2025. Extending those feeds back to January 2025 would add a 2025 line to every chart in this review, and is a short configuration change. Call tracking began in November 2025, so 2026 is its first full year. Facebook Ads detail exists from April 2026. Hours worked are not included here; the September brief carries the monthly hours and this review will carry the year's total once the January–September timesheets are added. Booked jobs, revenue, Kinetico's referred-lead counts, and radio and print spend remain outside the data set.
Sources: website lead forms, CallRail, Google Analytics, Google Ads, Google Search Console, Google Business Profile, Facebook, Instagram, SEMrush Position Tracking and AI topics. Data through Sep 30, 2026. Media Relations Agency · Haferman Water Conditioning · 2026 Year-to-Date Review