The Mid-Peninsula Doesn't Move as One Market, and the Price Data Proves It

Portola Valley's home values are up nearly 10 percent over the past year. They are also flat. Both numbers are current, both come from data through the summer of 2026, and both are describing the same town over the same twelve months.

That is not a typo and it is not a data error. It is what happens when a town has too few closings for a single number to mean what everyone assumes it means, and it is the reason a buyer comparing Atherton, Woodside, Portola Valley, and Menlo Park needs to know which of those four price stories can actually be trusted as a planning input and which ones are closer to noise dressed up as a trend.

Two numbers, one town, same twelve months

Zillow's automated valuation model, republished by InMenlo on July 15, 2026 using Stacker's statewide ranking of California's most expensive cities, put Portola Valley's typical home value at $4,253,425, a gain of 9.5 percent over the prior year. Stacker's list placed Portola Valley fourth among all California cities tracked, with Atherton first, Woodside 20th, and Menlo Park 21st.

Redfin's tracking, based on homes that actually closed escrow and updated through July 2026, tells a different story for the same town over roughly the same window. The median sale price over the three months ending May 2026 was $4.6 million, down about a tenth of a percent from the same period a year earlier. Price per square foot fell 10.6 percent year over year. Days on market rose from 14 to 18. Only 13 homes sold in Portola Valley in May 2026, the same count as May 2025.

One measure says values climbed nearly 10 percent. The other says the median barely moved and the price per square foot actually dropped. Both are real. They are just measuring different things.

Why this happens where sales are rare

Zillow's model estimates value across the entire housing stock, whether or not a given home sold this year. Redfin's median only reflects the homes that actually changed hands in a given window. In a market with hundreds of monthly closings, those two approaches tend to converge, because the closed sample is large enough to represent the whole stock. In a market where only 13 homes sold in a month, the closed sample is the whole story, and which 13 homes happened to sell can swing the median in either direction regardless of what the broader stock is doing.

MLSListings, the multiple listing service used by nearly all licensed Silicon Valley agents, publishes trend data for both Woodside and Portola Valley that, as of its last update on June 30, 2026, describes this exact pattern in its own commentary. Both towns show sale-to-list price ratios that alternate between overbids and discounts rather than moving consistently in one direction, meaning buyers are not facing uniform competitive pressure. Both show days on market that vary without any clear seasonal rhythm. Both are described as having a low number of sales with no consistent pattern, and outcomes that depend on the specific property rather than the overall market. The plain language version: in these two towns, there is no aggregate market in the way that phrase usually implies. There is a small handful of individual transactions each month, each one capable of moving the town-wide number on its own.

That also explains Woodside's more modest headline. Zillow's figure put Woodside's typical home value at $3,926,400, up 4.0 percent over the year, the smallest gain of the four towns. The same MLSListings data describes Woodside's price-per-square-foot trend as flat to slightly rising over the past three years, following the sharper 2022 spike. Read alongside the sale-to-list inconsistency, a 4 percent aggregate gain in Woodside is not evidence of a cooling market. It is evidence of a market too thin and too property-specific for a single percentage to describe.

Where the aggregate actually holds up

Atherton looks different. Zillow's figure put its typical home value at $8,220,143 in the July 2026 data, up 10.5 percent over the year, the largest gain among the four towns and the highest ranking on Stacker's statewide list. Palo Alto Online, reporting separately in April 2026 on actual market conditions, put Atherton's median home price at more than $7.4 million as of February 2026. Two independent sources, checked at two different points in the year, point in the same direction and land in a similar range. That kind of agreement across an automated model and reported actual pricing is the closest thing to a trustworthy signal in this comparison.

Part of the reason is deal flow. Compass reported 40 sales of homes priced at $5 million or above during February 2026 across the broader Midpeninsula, including Atherton, Woodside, Portola Valley, Menlo Park, Palo Alto, Los Altos, and Los Altos Hills, compared with 30 sales in that price range in February 2025. Susan Sims, a broker with The Agency Los Altos, described the year as opening with unusual strength even through what is typically a slower winter stretch. Alexander Lewicki, director of listings for DeLeon Realty, called the luxury segment stronger than anticipated. That same reporting noted a countercurrent: state WARN filings showed hundreds of Bay Area job cuts in early 2026, including in Menlo Park and neighboring cities, and some buyers were described as holding off to see how layoffs and broader economic uncertainty played out. The luxury tier kept moving. The more price-sensitive tier got more cautious. Both things were true in the same quarter, in the same region.

Even the winners carry an asterisk

None of this makes Atherton's weekly numbers immune to the same small-sample problem. A weekly sales report covering activity between June 16 and July 10, 2026, and made public in early August, recorded only three sales in Atherton that week, producing an $11 million median, well above the town's own annual figure. The same week, Los Altos, Woodside, and Portola Valley posted medians ranging from $3.95 million to $5.62 million. The single highest sale of that week was an 8-acre compound in Los Altos Hills that closed at $30 million. The property was originally built in 2004 by Brocade Communications Systems founder Kumar Malavalli, listed for $88 million as far back as 2016, and returned to market in May 2026 asking $36.9 million before closing at $30 million. One property, one closing, and it single-handedly defined that week's regional headline.

The lesson is not that Atherton's data is unreliable. It is that any number drawn from a single week or a single month, in any of these towns, tells you almost nothing on its own. The annual figures hold up better in Atherton because there is enough underlying activity across a full year to smooth out the week-to-week swings. In Woodside and Portola Valley, the same smoothing does not fully happen, because the annual total is still built from so few transactions that one unusual estate sale or one unusual bargain can carry outsized weight.

What this actually means if you're comparing these four towns

Menlo Park sits at a different scale than the other three. Its Zillow-based typical value was $2,869,009 in the July 2026 data, up 5.5 percent over the year, well short of Atherton's pace but still ahead of Woodside's 4.0 percent, and it sits at the most conventional price tier of the group. For a buyer weighing family-sized homes rather than architectural estates, Menlo Park's numbers are more likely to behave the way a normal market's numbers behave.

For Atherton, the annual direction is worth trusting. The week-to-week number is not.

For Woodside and Portola Valley, the town-wide figure is closer to a rough neighborhood description than a forecast. The more useful question is not "what is the market doing" but "what did the three or four most comparable properties actually sell for this year, and how long did they sit."

A few questions worth putting to whoever is guiding your search:

• Which specific closed sales in the last six months actually resemble the property I'm considering, in size, condition, and lot?
• Was the town's reported median this quarter shaped by one unusually large or unusually modest sale?
• How does the automated valuation for this property compare with the actual closed comps, and where do the two disagree?

A short FAQ

Why would an automated estimate and an actual closing price disagree for the same town? An automated model estimates value across every home in the town, whether or not it sold. A median sale price only reflects the homes that actually closed in that window. When very few homes close in a given period, the closed group can look nothing like the town as a whole, and the two numbers drift apart.

Does a strong five-year gain mean it's a good time to buy in a given town? A longer window smooths out some of the noise from any single quarter, but it still describes the past, not your specific transaction. Broader conditions, including the hiring and layoff patterns reported across the region in 2026, can shift faster than any five-year average reflects. This is not financial or investment advice, and the right read for your situation depends on the specific property and your own timeline.

The number quoted in a headline is rarely the number that should decide anything. What decides a Mid-Peninsula transaction is the handful of comparable sales that actually happened near the property in question, and how well someone can read the difference between a town where that comparison is stable and a town where it moves with every closing.

Michael Warren Real Estate works across Atherton, Woodside, Portola Valley, and Menlo Park with exactly this kind of property-by-property comparison, backed by Compass's data tools and direct knowledge of what's actually closing in each town. If you're weighing these four markets against each other, request a confidential consultation before you lean on any single headline number.